{ "name": "Is it just a wrapper?", "description": "Estimated inference cost, markup multiple and verdict for AI products. Estimates, not measurements: costs are what the stated usage would cost at published API rates, and real vendor costs are usually lower, so every multiple is a floor.", "url": "https://isitjustawrapper.com/", "methodology": "https://isitjustawrapper.com/methodology/", "corrections": "https://isitjustawrapper.com/corrections/", "license": "CC BY 4.0 — https://creativecommons.org/licenses/by/4.0/", "citation": "Is it just a wrapper? (isitjustawrapper.com)", "generatedAt": "2026-09-19T08:23:30.694Z", "verdicts": { "fair": "The price is mostly buying things that are not the model: evals, data pipelines, integrations, compliance, support, real interface work.", "steep": "Genuine work exists on top of the API, but the multiple is high for what it buys. A motivated user has cheaper paths available.", "wrapper": "An interface and a system prompt over a public API. This describes an architecture, not a vendor’s conduct: shipping one is legal, common, and sometimes exactly what a customer wants." }, "rateCard": { "updatedOn": "2026-08-11", "note": "Archetype rates, not vendor-specific quotes. Entries cite a tier of model ('mid-tier frontier text model') rather than a named product, because naming one vendor's model would make every estimate on the site wrong the day that vendor repriced. Each archetype is set at a representative published rate for its tier and links to a real rate card so the number can be audited. Costs are computed from this file at build time; no entry stores a cost figure of its own. When a rate here changes, every affected estimate on the site changes with it in one commit.", "models": { "frontier-large": { "label": "Large frontier text model", "note": "Top-tier reasoning model. Representative of the most expensive publicly listed general-purpose text APIs.", "units": { "inputTokens": { "usdPer": 15, "per": 1000000, "display": "input tokens" }, "outputTokens": { "usdPer": 75, "per": 1000000, "display": "output tokens" } }, "source": { "kind": "rate-card", "label": "Anthropic API pricing", "url": "https://claude.com/pricing" }, "checkedOn": "2026-08-11" }, "frontier-mid": { "label": "Mid-tier frontier text model", "note": "The default workhorse tier. Most products in this directory are estimated against this unless their task obviously needs more or less.", "units": { "inputTokens": { "usdPer": 3, "per": 1000000, "display": "input tokens" }, "outputTokens": { "usdPer": 15, "per": 1000000, "display": "output tokens" } }, "source": { "kind": "rate-card", "label": "Anthropic API pricing", "url": "https://claude.com/pricing" }, "checkedOn": "2026-08-11" }, "frontier-small": { "label": "Small fast text model", "note": "Cheap, low-latency tier used for autocomplete, classification and short-form generation.", "units": { "inputTokens": { "usdPer": 0.8, "per": 1000000, "display": "input tokens" }, "outputTokens": { "usdPer": 4, "per": 1000000, "display": "output tokens" } }, "source": { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" }, "checkedOn": "2026-08-11" }, "frontier-nano": { "label": "Nano text model", "note": "The cheapest hosted tier, appropriate for inline completion where latency matters more than depth.", "units": { "inputTokens": { "usdPer": 0.1, "per": 1000000, "display": "input tokens" }, "outputTokens": { "usdPer": 0.4, "per": 1000000, "display": "output tokens" } }, "source": { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" }, "checkedOn": "2026-08-11" }, "embedding-standard": { "label": "Text embedding model", "units": { "inputTokens": { "usdPer": 0.02, "per": 1000000, "display": "embedded tokens" } }, "source": { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" }, "checkedOn": "2026-08-11" }, "asr-standard": { "label": "Speech-to-text transcription", "note": "Hosted transcription billed per minute of audio. Self-hosting an open ASR model is cheaper again; see each entry's DIY path.", "units": { "audioMinutes": { "usdPer": 0.006, "per": 1, "display": "audio minutes", "displaySingular": "audio minute" } }, "source": { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" }, "checkedOn": "2026-08-11" }, "tts-standard": { "label": "Text-to-speech synthesis", "units": { "characters": { "usdPer": 15, "per": 1000000, "display": "synthesised characters" } }, "source": { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" }, "checkedOn": "2026-08-11" }, "image-gen-api": { "label": "Hosted image generation", "units": { "images": { "usdPer": 0.035, "per": 1, "display": "images", "displaySingular": "image" } }, "source": { "kind": "rate-card", "label": "Replicate pricing", "url": "https://replicate.com/pricing" }, "checkedOn": "2026-08-11" }, "gpu-hour-highend": { "label": "Rented high-end GPU", "note": "A100/H100-class rental, for video and music generation where a mid-range card is not a credible substitute. Used where a product trains its own generative media models and publishes no comparable API.", "units": { "gpuHours": { "usdPer": 2.5, "per": 1, "display": "GPU-hours", "displaySingular": "GPU-hour" } }, "source": { "kind": "rate-card", "label": "RunPod GPU pricing", "url": "https://www.runpod.io/pricing" }, "checkedOn": "2026-08-11" }, "gpu-hour-commodity": { "label": "Rented commodity GPU", "note": "Mid-range rental GPU (L4/A10 class) for running open weights yourself. Used where a product has no comparable public API and the honest alternative is self-hosting.", "units": { "gpuHours": { "usdPer": 0.6, "per": 1, "display": "GPU-hours", "displaySingular": "GPU-hour" } }, "source": { "kind": "rate-card", "label": "RunPod GPU pricing", "url": "https://www.runpod.io/pricing" }, "checkedOn": "2026-08-11" }, "web-search-api": { "label": "Web search API", "units": { "queries": { "usdPer": 6, "per": 1000, "display": "search queries" } }, "source": { "kind": "rate-card", "label": "Brave Search API pricing", "url": "https://brave.com/search/api/" }, "checkedOn": "2026-08-11" } } }, "entries": [ { "slug": "jasper-pro", "name": "Jasper Pro", "vendor": "Jasper AI", "category": "writing", "page": "https://isitjustawrapper.com/app/jasper-pro/", "productUrl": "https://www.jasper.ai", "tier": "Pro", "monthlyUsd": 69, "billingNote": "$69 per seat per month on monthly billing; $59 per seat per month on annual, which the page advertises as roughly 20% off. Business pricing is quoted rather than published. The earlier 'Creator' tier this entry was first written against no longer appears on the pricing page.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 1.08, "markupMultiple": 63.89, "verdict": "wrapper", "verdictSummary": "Marketing copy generated through templates over third-party models, at an estimated multiple in the sixties on typical usage.", "verdictReasoning": "The generation runs on models Jasper licenses rather than trains, and at this volume the estimated token bill is around a dollar. Template libraries, brand voice settings and campaign organisation are real features and some teams value them highly. The verdict reflects where the money goes rather than whether the software is any good: at this multiple, the interface and the workflow are what is being sold, and both are reproducible. A team buying several seats should note the multiple applies per seat.", "usageAssumption": "60 marketing pieces a month — blog sections, ad variants, product descriptions — at roughly 1,500 input and 900 output tokens each.", "valueAdd": [ "ux-craft", "integrations" ], "whatYouLose": "Templates you did not have to write, brand voice consistency across a team, and a workflow non-technical marketers can use without touching a prompt.", "diyPath": "A shared prompt library in any assistant subscription covers most of it. For a team, a small internal tool over an API key with saved brand-voice prompts reaches parity quickly.", "sources": [ { "kind": "pricing", "label": "Jasper pricing", "url": "https://www.jasper.ai/pricing" }, { "kind": "rate-card", "label": "Anthropic API pricing", "url": "https://claude.com/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11", "note": "Price basis standardised to monthly billing on 2026-08-11. The page was checked by @dovisvai on the same day; only which advertised rate the entry headlines has changed." }, "lastReviewed": "2026-08-11" }, { "slug": "wordtune-premium", "name": "Wordtune Premium", "vendor": "Wordtune (AI21 Labs)", "category": "writing", "page": "https://isitjustawrapper.com/app/wordtune-premium/", "productUrl": "https://www.wordtune.com", "tier": "Unlimited", "monthlyUsd": 19.99, "billingNote": "$19.99/month on the Unlimited tier, billed monthly. Annual billing is advertised at $9.99/month, which halves the multiple below. A free tier with a daily rewrite cap exists and is the comparison most readers should be making.", "priceCheckedOn": "2026-09-18", "estimatedApiCostUsd": 0.43, "markupMultiple": 46.27, "verdict": "wrapper", "verdictSummary": "Sentence rewriting delivered through a browser extension, where the unit of work is small enough that essentially the whole subscription is the interface rather than the generation.", "verdictReasoning": "The distinguishing feature of this category is how little inference a rewrite actually takes: a sentence in, a few sentences out, on a model that does not need to be large. That makes the gap between price and token cost wider here than almost anywhere else in the directory. The extension is genuinely convenient and the vendor has real language-model research behind it, but what a Premium subscriber is buying at the margin is the button in the text box.", "usageAssumption": "150 sentence rewrites a month — the tone and length adjustments the product is bought for — at roughly 600 input and 600 output tokens each, on a small fast model.", "valueAdd": [ "ux-craft", "distribution" ], "whatYouLose": "In-place rewriting inside any text field. For someone who rewrites constantly, removing the round trip is worth something real.", "diyPath": "The free tier covers light use. Beyond that, any assistant subscription rewrites sentences on request, and with an API key the estimated monthly token cost at this volume is well under a dollar.", "sources": [ { "kind": "pricing", "label": "Wordtune pricing", "url": "https://www.wordtune.com/pricing" }, { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-09-18" }, "lastReviewed": "2026-09-18" }, { "slug": "grammarly-pro", "name": "Grammarly Pro", "vendor": "Grammarly", "category": "writing", "page": "https://isitjustawrapper.com/app/grammarly-pro/", "productUrl": "https://www.grammarly.com", "tier": "Pro", "monthlyUsd": 30, "billingNote": "$30/member/month month-to-month. Grammarly also advertises $12/member/month on annual billing, which cuts the multiple below to roughly a sixth of what is shown. The free tier already covers most non-generative checking, which is the comparison most readers should be making.", "priceCheckedOn": "2026-09-18", "estimatedApiCostUsd": 0.75, "markupMultiple": 40, "verdict": "steep", "verdictSummary": "An estimated forty-fold multiple month-to-month, most of it for checking that the free tier largely already does.", "verdictReasoning": "Grammarly's models are its own and its integration surface is unusually wide, both of which are real work. The difficulty is the comparison a subscriber should actually make: not against an API, but against Grammarly's own free tier, which handles most of what most people use it for. A forty-fold multiple over estimated inference for the generative increment is a lot to pay for a narrow delta. Annual billing cuts it to around sixteen, and that is the version of this product that looks reasonable.", "usageAssumption": "40,000 words checked and 30 generative rewrites a month. Only the generative work is priced against a public API; the checking runs on Grammarly's own models and is given a separate allowance.", "valueAdd": [ "proprietary-model", "distribution", "ux-craft" ], "whatYouLose": "Checking inside every text field you touch without thinking about it. That ambient presence is not something a script reproduces.", "diyPath": "LanguageTool, self-hosted or on its own cheaper plan, covers the checking. For rewrites, any assistant with your own key. Neither follows you into every input box.", "sources": [ { "kind": "pricing", "label": "Grammarly plans", "url": "https://www.grammarly.com/plans" }, { "kind": "other", "label": "LanguageTool pricing (comparison)", "url": "https://languagetool.org/premium" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-09-18" }, "lastReviewed": "2026-09-18" }, { "slug": "rytr-unlimited", "name": "Rytr Unlimited", "vendor": "Rytr", "category": "writing", "page": "https://isitjustawrapper.com/app/rytr-unlimited/", "productUrl": "https://rytr.me", "tier": "Unlimited", "monthlyUsd": 9, "billingNote": "$9/month month-to-month. The pricing page's yearly toggle shows the same tier at $7.50/month equivalent, and Premium at $24.16/month. A free tier with a monthly character cap exists.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 0.25, "markupMultiple": 36.29, "verdict": "wrapper", "verdictSummary": "Short-form copy generated from a use-case menu over third-party models, at an estimated multiple in the mid-thirties on a low absolute price.", "verdictReasoning": "A high multiple on a small number: nine dollars against an estimated quarter of a dollar in tokens. The absolute price is low enough that the multiple may not matter to a buyer, and that is a legitimate way to feel about it. What the entry records is that essentially all of the price is the menu of use cases and the interface around it, not the generation.", "usageAssumption": "100 short pieces a month — captions, emails, product blurbs — at roughly 600 input and 500 output tokens each, priced against a small fast model.", "valueAdd": [ "ux-craft" ], "whatYouLose": "A curated list of use cases and a simple editor. For a non-technical user, the fifteen minutes of setup a self-built alternative needs may be the deciding cost.", "diyPath": "Any free assistant tier handles this volume of short-form copy. With an API key, the estimated monthly cost at this usage is well under a dollar.", "sources": [ { "kind": "pricing", "label": "Rytr pricing", "url": "https://rytr.me/pricing" }, { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11" }, "lastReviewed": "2026-08-11" }, { "slug": "midjourney-standard", "name": "Midjourney Standard", "vendor": "Midjourney", "category": "image", "page": "https://isitjustawrapper.com/app/midjourney-standard/", "productUrl": "https://www.midjourney.com", "tier": "Standard", "monthlyUsd": 30, "billingNote": "$30/month billed monthly, advertised lower on annual billing. Standard includes a fixed allowance of fast generation hours plus unmetered relaxed generation.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 0.84, "markupMultiple": 35.71, "verdict": "fair", "verdictSummary": "One of the highest multiples in this directory, on a product where the model itself is the thing being sold rather than a layer over someone else's.", "verdictReasoning": "This entry exists to make the site's central point. A multiple in the mid-thirties looks damning until you ask what the alternative produces: open weights on rented hardware are cheap and do not make Midjourney images. The house aesthetic is trained, not prompted, and there is no API to wrap even if you wanted to. A high multiple over a substitute that is not actually a substitute is a weak argument, and the verdict follows the argument rather than the number.", "usageAssumption": "600 images a month. Midjourney has no public API, so the comparison is the honest alternative: running open image weights on a rented GPU, at roughly 8 seconds per image.", "valueAdd": [ "proprietary-model", "scale-infra" ], "whatYouLose": "The aesthetic. Everything else about the product is replaceable; the output is not, and no amount of prompt engineering against open weights closes that gap.", "diyPath": "Flux or SDXL on a rented GPU, or ComfyUI locally if you have the hardware. Cheaper by a wide margin, more controllable, and it produces different pictures.", "sources": [ { "kind": "pricing", "label": "Midjourney subscription plans", "url": "https://www.midjourney.com/pricing" }, { "kind": "rate-card", "label": "RunPod GPU pricing", "url": "https://www.runpod.io/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11" }, "lastReviewed": "2026-08-11" }, { "slug": "intercom-fin", "name": "Intercom Fin", "vendor": "Intercom", "category": "support", "page": "https://isitjustawrapper.com/app/intercom-fin/", "productUrl": "https://www.intercom.com/fin", "tier": "Standalone, per resolution", "monthlyUsd": 297, "billingNote": "$0.99 per Fin outcome, running standalone against an existing helpdesk with no seats. The $297 here is NOT a published price: it is 300 outcomes at $0.99, so this row can sit in a table of monthly figures. Change the volume and the row changes with it. Fin is also bundled into Intercom seat plans at $29, $99 and $139 per seat.", "priceCheckedOn": "2026-09-18", "estimatedApiCostUsd": 9.51, "markupMultiple": 31.23, "verdict": "fair", "verdictSummary": "A high multiple over inference on a product that charges only when it actually resolves a conversation, and whose real competitor is a human support agent rather than an API call.", "verdictReasoning": "This is the entry that most tests the site's own method. Divide the price by the tokens and the multiple is large. But the comparison the arithmetic implies — that you could do this with an API key — is the wrong one twice over. First, the vendor absorbs every conversation Fin fails to resolve, so the price includes all the attempts you are not billed for. Second, the thing being replaced is a person answering a ticket, which costs many times $0.99 in any market. Outcome pricing also inverts the usual incentive: a vendor paid per resolution loses money when its model is bad, which is a stronger quality guarantee than any benchmark. The multiple is real and the verdict still goes the other way.", "usageAssumption": "300 conversations resolved a month. Each resolution is assumed to take several model turns over retrieved help-centre content — roughly 8,000 tokens of context and 500 of reply — plus keeping that content embedded for retrieval.", "valueAdd": [ "eval-harness", "integrations", "orchestration" ], "whatYouLose": "Billing that only charges for outcomes, the judgement about when to escalate to a human, and connectors into the helpdesk you already run. A self-built bot bills you for every failed attempt, because your API provider charges by the token regardless of whether the customer went away happy.", "diyPath": "A retrieval bot over your help centre with your own key is an afternoon's work and would cost a fraction of this per conversation. It will also answer confidently when it should escalate, and you pay for those answers too — which is the difference the price is charging for.", "sources": [ { "kind": "pricing", "label": "Intercom pricing", "url": "https://www.intercom.com/pricing" }, { "kind": "rate-card", "label": "Anthropic API pricing", "url": "https://claude.com/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-09-18" }, "lastReviewed": "2026-09-18" }, { "slug": "quillbot-premium", "name": "QuillBot Premium", "vendor": "QuillBot (Learneo)", "category": "writing", "page": "https://isitjustawrapper.com/app/quillbot-premium/", "productUrl": "https://quillbot.com", "tier": "Premium", "monthlyUsd": 19.95, "billingNote": "US list price, billed monthly. QuillBot prices by region — an automated check from outside the US returned Thai baht — so what a reader pays depends on where they are. The free tier paraphrases with a word limit and is the comparison most students should be making.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 0.77, "markupMultiple": 25.98, "verdict": "wrapper", "verdictSummary": "Paraphrasing and summarising at paragraph scale, where the operation is small, the model does not need to be large, and the interface is what the subscription buys.", "verdictReasoning": "Rewriting a paragraph is among the cheapest useful things a language model does, which puts a wide gap between this price and the tokens behind it. The product's real position is distribution rather than capability: it is the tool a very large number of students already have open, with modes and a word counter arranged for coursework. That is worth something, and it is not the model. Regional pricing complicates the comparison further — what a reader pays depends on where they are, which the multiple shown here cannot express.", "usageAssumption": "A heavy student user: 200 paraphrasing or summarising operations a month, each around 800 tokens in and 800 out, on a small fast model.", "valueAdd": [ "ux-craft", "distribution" ], "whatYouLose": "A tool shaped for coursework rather than a blank chat box, and the habit of already having it open. For its actual audience that framing matters more than capability does.", "diyPath": "The free tier covers occasional use. Any assistant subscription paraphrases on request, and at this volume an API key would cost well under a dollar a month in tokens.", "sources": [ { "kind": "pricing", "label": "QuillBot Premium", "url": "https://quillbot.com/premium" }, { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "copy-ai-chat", "name": "Copy.ai Chat", "vendor": "Copy.ai", "category": "writing", "page": "https://isitjustawrapper.com/app/copy-ai-chat/", "productUrl": "https://www.copy.ai", "tier": "Chat", "monthlyUsd": 29, "billingNote": "$29/month billed monthly; $24/month equivalent on annual billing at $288 up front. The self-serve tiers above this jump to $1,000/month and are aimed at go-to-market teams rather than individual writers, so they are not covered by this entry.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 1.25, "markupMultiple": 23.24, "verdict": "wrapper", "verdictSummary": "Marketing copy generated through saved workflows over third-party models, at an estimated multiple in the twenties on typical individual usage.", "verdictReasoning": "The generation runs on models Copy.ai licenses rather than trains, and at this volume the estimated token bill is a little over a dollar. The workflow builder is a real piece of software and the higher tiers sell something different — orchestration across a sales stack, at prices that imply a different buyer entirely. For the individual writer paying twenty-nine dollars, the interface and the saved prompts are what the money buys, and both are reproducible.", "usageAssumption": "80 marketing pieces a month — ad variants, product descriptions, email copy — at roughly 1,200 input and 800 output tokens each, generated on a mid-tier frontier model.", "valueAdd": [ "ux-craft", "orchestration" ], "whatYouLose": "Workflows you did not have to build, and a shared place for a team's prompts to live. For someone who does not want to think about prompting, that convenience is the product.", "diyPath": "A shared prompt library in any assistant subscription covers the individual case. For a team, a small internal tool over an API key reaches parity on the entry tier quickly, though not on the automation tiers above it.", "sources": [ { "kind": "pricing", "label": "Copy.ai pricing", "url": "https://www.copy.ai/prices" }, { "kind": "rate-card", "label": "Anthropic API pricing", "url": "https://claude.com/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "elevenlabs-creator", "name": "ElevenLabs Creator", "vendor": "ElevenLabs", "category": "audio", "page": "https://isitjustawrapper.com/app/elevenlabs-creator/", "productUrl": "https://elevenlabs.io", "tier": "Creator", "monthlyUsd": 22, "billingNote": "$22/month at list, with a discounted first month commonly advertised. Billed in credits rather than minutes, and overage is charged per credit.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 1.5, "markupMultiple": 14.67, "verdict": "fair", "verdictSummary": "An estimated mid-teens multiple over commodity speech synthesis, buying voice models that commodity synthesis does not sound like.", "verdictReasoning": "The comparison here is deliberately unflattering to the vendor: it prices the work against the cheapest hosted synthesis available, which does not sound the same. The weights are trained in-house and the output quality is the reason people buy, which is exactly the situation the proprietary-model tag exists for. Voice cloning also carries consent and misuse obligations that a raw API does not hand you.", "usageAssumption": "100,000 characters of synthesised speech a month — roughly two hours of finished audio — priced against a commodity hosted text-to-speech API.", "valueAdd": [ "proprietary-model", "scale-infra" ], "whatYouLose": "The specific voice quality, cloning from short samples, streaming latency, and the licensing framework around voices that makes commercial use straightforward.", "diyPath": "Open speech models such as XTTS or Piper run on a rented GPU for a fraction of this. They are genuinely usable, and they are also audibly not the same product.", "sources": [ { "kind": "pricing", "label": "ElevenLabs pricing", "url": "https://elevenlabs.io/pricing" }, { "kind": "rate-card", "label": "OpenAI API pricing (speech)", "url": "https://openai.com/api/pricing/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11" }, "lastReviewed": "2026-08-11" }, { "slug": "chatpdf-plus", "name": "ChatPDF Plus", "vendor": "ChatPDF", "category": "research", "page": "https://isitjustawrapper.com/app/chatpdf-plus/", "productUrl": "https://www.chatpdf.com", "tier": "Plus", "monthlyUsd": 19.99, "billingNote": "US list price, billed monthly; annual billing is advertised lower. ChatPDF prices regionally — the same Plus plan has been observed at roughly $9 equivalent in some markets (289 THB in Thailand), so what a reader pays depends on where they are. This entry uses the US price because that is what most readers of this site will be shown. A free tier with page and document caps exists.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 1.37, "markupMultiple": 14.54, "verdict": "wrapper", "verdictSummary": "Textbook retrieval over uploaded PDFs at an estimated multiple around fifteen, using techniques that are now standard and widely available free.", "verdictReasoning": "Chunk, embed, retrieve, answer. The pipeline is the one described in every retrieval tutorial published since 2023, and the same capability now ships inside assistant products people already pay for. The upload-and-ask interface is clean and the parsing handles awkward PDFs better than a naive script would. Neither observation moves this off the estimated multiple, because neither is what the fifteen-fold is buying.", "usageAssumption": "30 documents of about 40 pages each, embedded once on upload, then 200 questions answered over the retrieved passages rather than the whole document.", "valueAdd": [ "ux-craft" ], "whatYouLose": "Drag, drop, ask. No setup, no keys, no chunking decisions. For occasional use that convenience is the entire proposition.", "diyPath": "Upload the PDF directly to any major assistant — most now accept documents on free or existing paid tiers. For volume, a local retrieval tool such as AnythingLLM or Open WebUI does the same pipeline over your own key.", "sources": [ { "kind": "pricing", "label": "ChatPDF home page (plan description)", "url": "https://www.chatpdf.com" }, { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11", "note": "Restored to the US list price on 2026-08-11. Regional pricing variance is recorded in the billing note and remains an open question with the vendor." }, "lastReviewed": "2026-08-11" }, { "slug": "notion-ai", "name": "Notion AI", "vendor": "Notion Labs", "category": "productivity", "page": "https://isitjustawrapper.com/app/notion-ai/", "productUrl": "https://www.notion.com/product/ai", "tier": "AI add-on (per member)", "monthlyUsd": 10, "billingNote": "Sold as a per-member add-on on top of an existing Notion plan. Notion has repackaged AI between add-on and bundled pricing more than once, so this entry goes stale faster than most — the live pricing page is the authority, not this row.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 0.9, "markupMultiple": 11.08, "verdict": "steep", "verdictSummary": "An estimated multiple around ten on light use, for features that are convenient where your documents already live but are not otherwise distinctive.", "verdictReasoning": "The retrieval work over a live workspace is real and the permission-aware search behind it is harder than it looks. The multiple is still high for what the average member uses, and the honest description of the value is that the AI is where the documents are. Heavy users move this toward fair; the seat-based pricing means the average member is not a heavy user.", "usageAssumption": "A realistic rather than heroic user: 25 AI actions a month — summaries, rewrites and workspace questions — plus keeping the workspace embedded for retrieval.", "valueAdd": [ "distribution", "integrations", "ux-craft" ], "whatYouLose": "Answers scoped correctly to your permissions, and zero setup. Any self-built alternative starts by solving the access-control problem, which is the part people underestimate.", "diyPath": "Export the workspace via the Notion API into any retrieval tool and query it with your own key. You will rebuild the permission model yourself, and that is the whole job.", "sources": [ { "kind": "pricing", "label": "Notion pricing", "url": "https://www.notion.com/pricing" }, { "kind": "docs", "label": "Notion AI product page", "url": "https://www.notion.com/product/ai" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11" }, "lastReviewed": "2026-08-11" }, { "slug": "chatbase-hobby", "name": "Chatbase Hobby", "vendor": "Chatbase", "category": "support", "page": "https://isitjustawrapper.com/app/chatbase-hobby/", "productUrl": "https://www.chatbase.co", "tier": "Hobby", "monthlyUsd": 40, "billingNote": "Billed monthly. Tiers are bounded by message credits and the number of chatbots rather than by seats, and messages beyond the allowance are billed on top of the subscription.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 6.43, "markupMultiple": 6.22, "verdict": "wrapper", "verdictSummary": "Retrieval over your own documents, wired to a chat widget, using a pipeline that has been standard since 2023 and is available free in several open tools.", "verdictReasoning": "Point it at a website, chunk, embed, retrieve, answer. The hosting, the widget and the ingestion of awkward source formats are real work, and for a small business with no engineer they are the entire reason to pay. What the price is not buying is anything about the answering itself, which runs on a licensed model doing the most documented task in the field. Compare this against Intercom Fin in the same category: one charges for a bot, the other charges only when a conversation is actually resolved, and that difference is the whole argument.", "usageAssumption": "A small business support bot answering 1,000 messages a month over its own documentation: each answer retrieves a few passages, sends roughly 6,000 tokens of context, and returns 400.", "valueAdd": [ "ux-craft", "integrations" ], "whatYouLose": "Hosting, the embed widget, and never thinking about chunking. Someone has to run the thing, and if it is not the vendor it is you.", "diyPath": "Open tools such as AnythingLLM, Flowise or Open WebUI do the same pipeline over your own key, self-hosted. Cheaper per message by a wide margin, and you own the uptime.", "sources": [ { "kind": "pricing", "label": "Chatbase pricing", "url": "https://www.chatbase.co/pricing" }, { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "freed", "name": "Freed", "vendor": "Freed AI", "category": "healthcare", "page": "https://isitjustawrapper.com/app/freed/", "productUrl": "https://www.getfreed.ai", "tier": "Core", "monthlyUsd": 79, "billingNote": "Core tier, per clinician per month on monthly billing. The page also lists Starter at $39 and Premier at $119 monthly ($104 annual), and shows promotional strikethrough pricing, so what counts as list price moves. Group and health-system pricing is quoted rather than published.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 12.96, "markupMultiple": 6.1, "verdict": "fair", "verdictSummary": "An estimated single-digit multiple over inference, against a product whose real costs are the regulated environment it operates in rather than the tokens.", "verdictReasoning": "This is the clearest case on the site for why the multiple does not decide the verdict. A clinical scribe carries a BAA, audit logging, data residency commitments, retention policy and the liability that follows a note entering a patient record. Those obligations cost money whether or not a single token is generated, and none of them appear in a token bill. The multiple would still be defensible at several times this figure.", "usageAssumption": "A clinician seeing 80 patients a month at about 18 minutes each: 1,440 minutes of consultation audio transcribed, then a structured clinical note generated per visit.", "valueAdd": [ "compliance-regulatory", "eval-harness", "ux-craft" ], "whatYouLose": "The BAA and the compliance posture behind it. Self-hosting this for a practice means you personally own HIPAA exposure, breach notification duties, and whatever a wrong note does to a patient record.", "diyPath": "Whisper self-hosted plus a note-generation prompt is technically straightforward and genuinely cheap. It is also the option that puts you, not a vendor, on the wrong end of a HIPAA audit. Cost is not the binding constraint for this category.", "sources": [ { "kind": "pricing", "label": "Freed pricing", "url": "https://www.getfreed.ai/pricing" }, { "kind": "docs", "label": "Freed security and compliance", "url": "https://www.getfreed.ai/security" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11" }, "lastReviewed": "2026-08-11" }, { "slug": "otter-ai-pro", "name": "Otter.ai Pro", "vendor": "Otter.ai", "category": "meetings", "page": "https://isitjustawrapper.com/app/otter-ai-pro/", "productUrl": "https://otter.ai", "tier": "Pro", "monthlyUsd": 16.99, "billingNote": "$16.99/user/month billed monthly; the annual rate is advertised at roughly half that, which roughly halves the multiple shown here. A free tier with a lower monthly minute cap exists.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 3.28, "markupMultiple": 5.18, "verdict": "steep", "verdictSummary": "An estimated five-fold multiple over transcription and summarisation, in the category where the do-it-yourself path is unusually easy to walk.", "verdictReasoning": "Joining calendars, sitting in meetings as a bot and separating speakers reliably is more work than a weekend, and the calendar integrations are a real maintenance burden. Transcription is also the category where open models got good enough fastest: self-hosted Whisper is genuinely competitive, which makes a five-fold multiple harder to defend here than the same multiple would be elsewhere. On annual billing this moves close to fair.", "usageAssumption": "400 minutes of meetings transcribed a month — about five hours of calls a week — plus 15 generated summaries with action items.", "valueAdd": [ "integrations", "ux-craft", "scale-infra" ], "whatYouLose": "Automatic joining and recording without anyone remembering to press a button, plus the shared searchable archive your colleagues also use.", "diyPath": "Whisper self-hosted or via API over a local recording, then any model for summaries. Cheap and accurate. You lose the calendar automation and the bot, and you will notice.", "sources": [ { "kind": "pricing", "label": "Otter.ai pricing", "url": "https://otter.ai/pricing" }, { "kind": "rate-card", "label": "OpenAI API pricing (transcription)", "url": "https://openai.com/api/pricing/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11", "note": "Price basis standardised to monthly billing on 2026-08-11. The page was checked by @dovisvai on the same day; only which advertised rate the entry headlines has changed." }, "lastReviewed": "2026-08-11" }, { "slug": "gamma-plus", "name": "Gamma Plus", "vendor": "Gamma", "category": "productivity", "page": "https://isitjustawrapper.com/app/gamma-plus/", "productUrl": "https://gamma.app", "tier": "Plus", "monthlyUsd": 12, "billingNote": "$12/month month-to-month. Gamma sells credits rather than seats at the lower tiers, and annual billing is advertised meaningfully lower. The pricing page renders client-side, so this figure cannot be machine-checked and is confirmed by hand.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 2.9, "markupMultiple": 4.14, "verdict": "steep", "verdictSummary": "Deck generation where the layout engine is the genuine achievement and the model contribution is a modest amount of text and a few images.", "verdictReasoning": "The hard part of Gamma is not writing slide copy — it is making generated content land on a page that looks designed, at any length, without a human nudging boxes. That layout engine is real software and it is why the product is used as widely as it is. The model layer on top of it is thin and runs on licensed models, and the alternatives have closed considerably now that general assistants export presentations. Steep rather than wrapper because the thing you are paying for demonstrably is not the generation; it is just that the generation is what the marketing emphasises.", "usageAssumption": "20 decks a month generated from a prompt or an outline: text for a dozen slides each, plus a couple of generated images per deck.", "valueAdd": [ "ux-craft", "distribution" ], "whatYouLose": "Decks that look composed without opening a design tool, and a format that works as a link rather than a file. Reproducing that with a general assistant means accepting worse layout.", "diyPath": "A general assistant can draft the content and export to slides; the result is less polished. For anyone who wanted a document rather than a deck, that gap may not matter.", "sources": [ { "kind": "pricing", "label": "Gamma pricing", "url": "https://gamma.app/pricing" }, { "kind": "rate-card", "label": "Replicate pricing", "url": "https://replicate.com/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "perplexity-pro", "name": "Perplexity Pro", "vendor": "Perplexity AI", "category": "search", "page": "https://isitjustawrapper.com/app/perplexity-pro/", "productUrl": "https://www.perplexity.ai", "tier": "Pro", "monthlyUsd": 20, "billingNote": "$20/month billed monthly; an annual rate is advertised lower. Pro is frequently bundled free for a year through carrier and hardware promotions, which many subscribers are on.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 5.25, "markupMultiple": 3.81, "verdict": "fair", "verdictSummary": "A grounded answer costs real money on both sides — the search call and the long retrieved context — and the estimated multiple stays in low single digits.", "verdictReasoning": "Retrieval-augmented search is one of the few consumer AI products where the non-model costs are genuinely large: every question pays for crawling or a search API before a token is generated, and the retrieved context makes the input side of the model bill unusually heavy. A single-digit multiple over that combined cost is not where the site's attention belongs.", "usageAssumption": "A regular research user: 100 searches a month, each running one web search and then answering over roughly 12,000 tokens of retrieved page text.", "valueAdd": [ "data-pipeline", "orchestration", "ux-craft" ], "whatYouLose": "Citations that line up with the sentences they support, a retrieval layer tuned over a large query volume, and freshness on fast-moving topics.", "diyPath": "A search API key plus a model call gets you a working clone in an afternoon. Getting citations to actually match the claims is the part that takes months.", "sources": [ { "kind": "pricing", "label": "Perplexity Pro", "url": "https://www.perplexity.ai/pro" }, { "kind": "rate-card", "label": "Brave Search API pricing", "url": "https://brave.com/search/api/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11", "note": "Price basis standardised to monthly billing on 2026-08-11. The page was checked by @dovisvai on the same day; only which advertised rate the entry headlines has changed." }, "lastReviewed": "2026-08-11" }, { "slug": "heygen-creator", "name": "HeyGen Creator", "vendor": "HeyGen", "category": "video", "page": "https://isitjustawrapper.com/app/heygen-creator/", "productUrl": "https://www.heygen.com", "tier": "Creator", "monthlyUsd": 29, "billingNote": "$29/month billed monthly; the page offers annual billing at a lower effective rate without printing the figure. Pro is $49 and Business $149 monthly. Tiers are bounded by video minutes and avatar slots rather than by seats.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 7.91, "markupMultiple": 3.67, "verdict": "fair", "verdictSummary": "A low single-digit multiple over rented hardware, for avatar models trained in-house that the open alternatives visibly do not match.", "verdictReasoning": "Talking-head video is the category where the gap between a product and its open substitute is easiest to see: open lip-sync models exist and produce output that reads as uncanny within a second. HeyGen trains its own avatar models and carries the consent and likeness infrastructure that selling a person's face requires, which is a legal obligation rather than a compute cost and appears nowhere in a GPU bill. The pressure on this price is that avatar quality is improving fast in the open, not that the markup is unexplained.", "usageAssumption": "30 minutes of finished avatar video a month: a script synthesised to speech, then a trained avatar rendered and lip-synced to it. HeyGen publishes no comparable API for its own avatar models, so the comparison is assembling the same thing from commodity speech synthesis and open video weights on rented hardware.", "valueAdd": [ "proprietary-model", "compliance-regulatory", "scale-infra" ], "whatYouLose": "Avatar quality that does not read as synthetic, and the consent framework that makes using a person's likeness commercially defensible. The second one matters more than it sounds.", "diyPath": "Open lip-sync models such as the Wav2Lip or LatentSync families over a commodity speech API, on a rented GPU. Substantially cheaper, visibly rougher, and the likeness permissions become entirely your problem.", "sources": [ { "kind": "pricing", "label": "HeyGen pricing", "url": "https://www.heygen.com/pricing" }, { "kind": "rate-card", "label": "RunPod GPU pricing", "url": "https://www.runpod.io/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "fireflies-pro", "name": "Fireflies.ai Pro", "vendor": "Fireflies.ai", "category": "meetings", "page": "https://isitjustawrapper.com/app/fireflies-pro/", "productUrl": "https://fireflies.ai", "tier": "Pro", "monthlyUsd": 18, "billingNote": "$18 per seat per month month-to-month. The same seat is advertised at $10 on annual billing, which roughly halves the multiple below. Business and Enterprise sit above at $19 and $39 per seat on annual billing.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 5.06, "markupMultiple": 3.56, "verdict": "steep", "verdictSummary": "Meeting transcription and summaries in the category where the do-it-yourself path is unusually accessible, at a multiple that annual billing roughly halves.", "verdictReasoning": "This entry exists to test whether the verdict given to Otter generalises, and it does. Joining calls reliably across several platforms and keeping calendars connected is real, unglamorous maintenance. But transcription is the category where open models caught up fastest — self-hosted Whisper is genuinely competitive — which makes the same multiple harder to defend here than it would be somewhere the alternative is worse. On annual billing this moves toward fair, which is worth stating plainly rather than burying.", "usageAssumption": "Ten hours of meetings a month transcribed with speaker separation, plus 25 generated summaries with action items.", "valueAdd": [ "integrations", "scale-infra", "ux-craft" ], "whatYouLose": "Automatic joining and recording, and the shared searchable archive. Colleagues using the same archive is a network effect a personal script cannot reproduce.", "diyPath": "Whisper self-hosted or via API over a local recording, then any model for summaries. Cheap and accurate; you lose the calendar automation and the bot, and you will notice both.", "sources": [ { "kind": "pricing", "label": "Fireflies.ai pricing", "url": "https://fireflies.ai/pricing" }, { "kind": "rate-card", "label": "OpenAI API pricing (transcription)", "url": "https://openai.com/api/pricing/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "suno-pro", "name": "Suno Pro", "vendor": "Suno", "category": "audio", "page": "https://isitjustawrapper.com/app/suno-pro/", "productUrl": "https://suno.com", "tier": "Pro", "monthlyUsd": 8, "billingNote": "$8/month billed monthly, with an annual option the page advertises as saving $24 a year. Premier is $24/month. Pro includes a monthly credit allowance and grants a general commercial-use licence, which the free tier does not.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 2.63, "markupMultiple": 3.05, "verdict": "fair", "verdictSummary": "A low single-digit multiple over rented hardware, for music models trained in-house that open weights do not match, plus a commercial licence the cheaper path does not grant.", "verdictReasoning": "The models are the product and there is no API to wrap. What makes this entry unusual is that the subscription also buys a licence: the paid tier grants commercial use of what you generate, which is a legal position rather than a compute cost and does not appear anywhere in a GPU bill. Open music models exist and are improving quickly, which is the honest pressure on this price.", "usageAssumption": "150 generated tracks a month of around two minutes each. Suno publishes no API for its own models, so the comparison is running open music weights on rented high-end GPUs.", "valueAdd": [ "proprietary-model", "scale-infra" ], "whatYouLose": "The output quality, the coherent song structure, and the commercial licence. Generating something you may legally sell is not the same as generating something.", "diyPath": "Open music models such as the ACE-Step or YuE families on rented GPUs, or MusicGen for shorter instrumental pieces. Cheaper, rougher, and the licensing of what you produce becomes your problem to research.", "sources": [ { "kind": "pricing", "label": "Suno pricing", "url": "https://suno.com/pricing" }, { "kind": "rate-card", "label": "RunPod GPU pricing", "url": "https://www.runpod.io/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "descript-creator", "name": "Descript Creator", "vendor": "Descript", "category": "video", "page": "https://isitjustawrapper.com/app/descript-creator/", "productUrl": "https://www.descript.com", "tier": "Creator", "monthlyUsd": 35, "billingNote": "$35/month month-to-month; the same tier is advertised at $24/month on annual billing, which cuts the multiple below by roughly a third. Hobbyist and Business tiers sit either side. The page's monthly/annual toggle makes it easy to read the wrong figure — this one is the month-to-month price.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 12.52, "markupMultiple": 2.8, "verdict": "fair", "verdictSummary": "A low single-digit multiple once rendering and storage are counted, on a product where the editing model rather than the inference is the thing being sold.", "verdictReasoning": "Transcription is cheap and getting cheaper, which makes a naive reading of this look worse than it is. The product is not transcription: it is editing video by editing text, and the engineering that keeps a multitrack timeline synchronised with an edited transcript is the actual work. Storage and rendering also cost real money continuously. The honest pressure on this price is that the transcription half is now nearly free, and that gap will widen.", "usageAssumption": "A working podcaster or video editor: 20 hours of audio transcribed a month, 40 generative edits or overdubs, plus the rendering and storage that a video tool carries whether or not a model runs.", "valueAdd": [ "ux-craft", "scale-infra", "integrations" ], "whatYouLose": "Editing video by deleting words. Every cheaper path gives you a transcript and leaves you to cut the timeline yourself, which is the labour the product exists to remove.", "diyPath": "Whisper for transcription, then a conventional editor. Costs a fraction of this in compute and considerably more of your time, which for a working editor is the expensive resource.", "sources": [ { "kind": "pricing", "label": "Descript pricing", "url": "https://www.descript.com/pricing" }, { "kind": "rate-card", "label": "OpenAI API pricing (transcription)", "url": "https://openai.com/api/pricing/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "monica-pro", "name": "Monica Pro", "vendor": "Monica", "category": "assistant", "page": "https://isitjustawrapper.com/app/monica-pro/", "productUrl": "https://monica.im", "tier": "Pro", "monthlyUsd": 8.3, "billingNote": "$8.30/month. Tiers are bounded by a monthly quota of 'advanced' queries against frontier models, with cheaper models unmetered, so the quota matters more than the fee. Annual billing is advertised lower.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 3.15, "markupMultiple": 2.63, "verdict": "wrapper", "verdictSummary": "A browser extension that forwards your question to models it licenses from other companies, priced above what those companies charge for the same calls.", "verdictReasoning": "This is the clearest architecture in the directory: there is no model here, only routing. The extension is genuinely well made — summarising the page you are on, without a copy-paste round trip, is a real convenience — and offering several vendors' models behind one subscription has value for someone who does not want several subscriptions. But the thing being sold sits entirely between the user and someone else's API, which is what the word describes. Note that the same convenience argument defends it: for a non-technical user, an extension that works everywhere is worth more than a cheaper key they would never configure.", "usageAssumption": "150 substantive queries a month against a frontier model — the quota the paid tier exists to sell — each carrying roughly 4,000 tokens of page or document context and returning 600.", "valueAdd": [ "ux-craft", "distribution", "orchestration" ], "whatYouLose": "A model in the sidebar of every page you visit, and not having to choose which vendor to subscribe to. For someone who would otherwise pay for two assistants, the bundle may genuinely be cheaper.", "diyPath": "Any first-party assistant subscription, or an open browser extension such as Sider's free tier or a self-hosted equivalent pointed at your own API key. The key costs what the tokens cost.", "sources": [ { "kind": "pricing", "label": "Monica pricing", "url": "https://monica.im/pricing" }, { "kind": "rate-card", "label": "Anthropic API pricing", "url": "https://claude.com/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "runway-standard", "name": "Runway Standard", "vendor": "Runway", "category": "video", "page": "https://isitjustawrapper.com/app/runway-standard/", "productUrl": "https://runway.com", "tier": "Standard", "monthlyUsd": 15, "billingNote": "$15/month billed monthly; $12/month equivalent on annual billing, which the page advertises as 20% off. Standard includes 625 credits a month; Pro is $35 monthly and Max $95 monthly. Credits buy generation time, so heavy months need top-ups.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 6, "markupMultiple": 2.5, "verdict": "fair", "verdictSummary": "A low single-digit multiple over the rented hardware an equivalent open model would need, on a product whose own video models are the reason to buy it.", "verdictReasoning": "Video generation is the most compute-hungry category in this directory, and it is priced here against high-end GPU rental rather than the cheap mid-range card used elsewhere, because a weaker card is not a credible substitute. The models are trained in-house, so the same argument that defends other model-owners applies: a multiple over a substitute that does not produce the same output is a weak argument. Credits also make the price behave like metered billing once you exceed the allowance, which the sticker price hides.", "usageAssumption": "The included 625 credits, roughly two minutes of finished generated video. Runway publishes no comparable API for its own models, so the comparison is the honest alternative: running open video weights on rented high-end GPUs.", "valueAdd": [ "proprietary-model", "scale-infra", "ux-craft" ], "whatYouLose": "The specific models, the queue that returns a clip in minutes rather than when your rented instance gets round to it, and the editing tools wrapped around generation.", "diyPath": "Open video weights such as the Wan or LTX families on a rented A100, or ComfyUI locally with enough VRAM. Genuinely cheaper per second, substantially more setup, and different output.", "sources": [ { "kind": "pricing", "label": "Runway pricing", "url": "https://runway.com/pricing" }, { "kind": "rate-card", "label": "RunPod GPU pricing", "url": "https://www.runpod.io/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "higgsfield", "name": "Higgsfield", "vendor": "Higgsfield AI", "category": "video", "page": "https://isitjustawrapper.com/app/higgsfield/", "productUrl": "https://higgsfield.ai", "tier": "Starter", "monthlyUsd": 15, "billingNote": "Starter tier, billed monthly. Higgsfield sells credit-bounded tiers and runs frequent promotional pricing, so what counts as list price moves. Credits buy generation time, so a heavy month needs top-ups on top of the subscription.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 8.75, "markupMultiple": 1.71, "verdict": "fair", "verdictSummary": "A single-digit multiple over the high-end hardware an open equivalent would need, on models trained in-house for camera motion that general video models do not reproduce.", "verdictReasoning": "The product's claim is control rather than raw generation — specified camera moves and VFX behaviour that a general text-to-video model will not follow reliably. That is trained capability, not prompting, and there is no API for it to wrap. Video is also the most compute-hungry category here, which compresses the multiple before any argument is made. The honest pressure is that this is a fast-moving field where an open model absorbing the same capability would collapse the gap quickly.", "usageAssumption": "Around three minutes of finished generated video a month with camera-motion control, priced against running open video weights on rented high-end GPUs.", "valueAdd": [ "proprietary-model", "scale-infra", "ux-craft" ], "whatYouLose": "The motion control, and results in minutes rather than whenever a rented instance finishes. Open video models generate; they do not take direction as reliably.", "diyPath": "Open video weights such as the Wan or LTX families on a rented A100, driven by ControlNet-style conditioning for camera movement. Cheaper per second, considerably more setup, and less predictable direction.", "sources": [ { "kind": "pricing", "label": "Higgsfield pricing", "url": "https://higgsfield.ai/pricing" }, { "kind": "rate-card", "label": "RunPod GPU pricing", "url": "https://www.runpod.io/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "lovable-pro", "name": "Lovable Pro", "vendor": "Lovable", "category": "coding", "page": "https://isitjustawrapper.com/app/lovable-pro/", "productUrl": "https://lovable.dev", "tier": "Pro", "monthlyUsd": 25, "billingNote": "$25/month. Lovable meters usage in credits and heavy building exhausts a monthly allowance quickly, so the sticker price understates what an active user actually spends. The pricing page renders client-side and cannot be machine-checked.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 18, "markupMultiple": 1.39, "verdict": "fair", "verdictSummary": "Prompt-to-application building where each turn resends a whole project as context, leaving very little room between the subscription and its own token bill.", "verdictReasoning": "Agentic building has an unusual cost shape: the model must see the application to change it, so every turn is enormous on the input side, and a month of active use approaches or exceeds the fee. That is why the plans are credit-metered rather than flat — the vendor cannot afford flat pricing at this cost structure, which is itself evidence about where the money goes. The orchestration around the model, applying multi-file edits without corrupting a working project, is the engineering; the deployment and hosting attached to it are further cost this estimate ignores.", "usageAssumption": "100 build messages a month inside the included allowance. Each one sends the working state of an application — roughly 40,000 tokens — and returns around 4,000 tokens of edits across several files.", "valueAdd": [ "orchestration", "scale-infra", "ux-craft" ], "whatYouLose": "Building without a local environment, and a deployed URL at the end of it. For someone who does not have a toolchain, that is the entire proposition rather than a convenience.", "diyPath": "An agentic coding tool such as Aider or Cline against your own API key, plus your own hosting. Comparable capability, no credit ceiling, and you own the setup and the deployment.", "sources": [ { "kind": "pricing", "label": "Lovable pricing", "url": "https://lovable.dev/pricing" }, { "kind": "rate-card", "label": "Anthropic API pricing", "url": "https://claude.com/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "github-copilot", "name": "GitHub Copilot", "vendor": "GitHub (Microsoft)", "category": "coding", "page": "https://isitjustawrapper.com/app/github-copilot/", "productUrl": "https://github.com/features/copilot", "tier": "Pro (individual)", "monthlyUsd": 10, "billingNote": "Individual Pro plan billed monthly; an annual rate is advertised lower. Free, Business and Enterprise tiers are priced separately and are not covered by this entry.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 10.26, "markupMultiple": 0.97, "verdict": "fair", "verdictSummary": "At this usage the subscription is priced at roughly what the inference alone is estimated to cost, before any of the editor integration work.", "verdictReasoning": "The estimated inference cost lands within a few percent of the sticker price, which means essentially none of the $10 is markup on tokens at this volume. Whatever the plan is buying — the editor extensions, the context-gathering that makes completions relevant, the request routing — is not being funded by a margin on the model. Lighter users subsidise heavier ones, which is how a flat fee works, but the flat fee itself is not the interesting number here.", "usageAssumption": "A developer using it most working days: about 4,000 inline completions and 150 chat turns a month. Completions are priced against a small fast model, chat against a mid-tier one.", "valueAdd": [ "integrations", "scale-infra", "distribution" ], "whatYouLose": "Sub-second inline completion inside your editor, the context selection that makes those completions relevant, and extensions maintained against several IDEs that each break on their own schedule.", "diyPath": "Continue.dev or Tabby wired to an API key, or a local model through Ollama for completion-only use. Both are real options and both leave you owning the context-gathering and the latency tuning.", "sources": [ { "kind": "pricing", "label": "GitHub Copilot plans", "url": "https://github.com/features/copilot/plans" }, { "kind": "docs", "label": "GitHub Copilot documentation", "url": "https://docs.github.com/en/copilot" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11" }, "lastReviewed": "2026-08-11" }, { "slug": "cursor", "name": "Cursor", "vendor": "Anysphere", "category": "coding", "page": "https://isitjustawrapper.com/app/cursor/", "productUrl": "https://cursor.com", "tier": "Pro", "monthlyUsd": 20, "billingNote": "Pro includes a monthly allowance of model requests; use beyond the allowance is billed as usage-based overage on top of the $20. This estimate assumes a user staying inside the included allowance, which is the only comparison the sticker price supports.", "priceCheckedOn": "2026-08-11", "estimatedApiCostUsd": 27.98, "markupMultiple": 0.71, "verdict": "fair", "verdictSummary": "The estimated inference cost of a daily user exceeds the subscription price, so the plan is priced below its own token bill at this volume.", "verdictReasoning": "A markup below 1× means the arithmetic this site performs stops being the interesting question: at the assumed usage the vendor is not marking up inference, it is absorbing part of it and recovering the difference from lighter users and from usage-based overages. Read the multiple as a statement about the included allowance, not about the company's economics overall — the overage pricing is where heavy users actually pay, and this entry does not model that.", "usageAssumption": "A daily user working inside the included allowance: about 250 agent requests a month carrying substantial repository context, plus 8,000 tab completions on a cheap fast model.", "valueAdd": [ "orchestration", "ux-craft", "scale-infra" ], "whatYouLose": "Repository indexing you did not have to build, apply-and-review UX for multi-file diffs, and the latency budget that makes tab completion usable rather than annoying.", "diyPath": "Aider or Cline with your own API key gets close on the agent loop and costs what the tokens cost. At the assumed volume that is estimated to be more than the subscription, not less.", "sources": [ { "kind": "pricing", "label": "Cursor pricing", "url": "https://cursor.com/pricing" }, { "kind": "docs", "label": "Cursor documentation", "url": "https://docs.cursor.com" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-11" }, "lastReviewed": "2026-08-11" }, { "slug": "windsurf-pro", "name": "Windsurf", "vendor": "Cognition", "category": "coding", "page": "https://isitjustawrapper.com/app/windsurf-pro/", "productUrl": "https://windsurf.com", "tier": "Pro", "monthlyUsd": 20, "billingNote": "Pro is $20/mo month-to-month as verified 2026-09-18. Paid plans can buy extra usage at API pricing beyond included quotas; this estimate assumes a user staying inside the included Pro allowance.", "priceCheckedOn": "2026-09-18", "estimatedApiCostUsd": 27.98, "markupMultiple": 0.71, "verdict": "fair", "verdictSummary": "At Cursor-comparable daily usage the estimated inference cost exceeds the $20 Pro sticker, so the plan is priced below its own token bill before counting the IDE and agent product work.", "verdictReasoning": "Windsurf Pro sits in the same coding-agent band as Cursor: an IDE surface, Cascade/agent loops, repo context, and unlimited Tab-style edits rather than a chat box over a public API. Matching the Cursor usage assumption keeps the multiple comparable. A markup below 1x means the interesting product is the orchestration and editor work, not a margin on tokens inside the included allowance. Heavy users who buy overages pay elsewhere; this entry does not model overage.", "usageAssumption": "A daily user working inside the included Pro allowance, aligned with the Cursor Pro estimate: about 250 agent requests a month carrying substantial repository context, plus 8,000 tab completions on a cheap fast model.", "valueAdd": [ "orchestration", "ux-craft", "scale-infra" ], "whatYouLose": "An agent-aware IDE with maintained Cascade flows, tab completion latency budget, and cloud agent integration you would otherwise wire yourself.", "diyPath": "Aider, Cline, or Continue with your own API key. At the assumed volume the token bill alone is estimated above $20, before the IDE work.", "sources": [ { "kind": "pricing", "label": "Devin / Windsurf pricing", "url": "https://devin.ai/pricing" }, { "kind": "docs", "label": "Windsurf site", "url": "https://windsurf.com" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-09-18" }, "lastReviewed": "2026-09-18" }, { "slug": "adobe-firefly", "name": "Adobe Firefly", "vendor": "Adobe", "category": "image", "page": "https://isitjustawrapper.com/app/adobe-firefly/", "productUrl": "https://www.adobe.com/products/firefly.html", "tier": "Firefly Standard", "monthlyUsd": 9.99, "billingNote": "Firefly Standard, billed monthly. Sold both standalone and bundled into Creative Cloud subscriptions, and tiers are bounded by generative credits rather than seats, so what a credit buys matters as much as the fee. The plans page defeats automated checks, so this figure is confirmed by hand.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 14, "markupMultiple": 0.71, "verdict": "fair", "verdictSummary": "Priced at or below what the same volume of generation costs on a commodity API, on a model trained specifically so its output can be used commercially without inheriting a copyright argument.", "verdictReasoning": "The distinguishing feature is not the pictures, it is the provenance. Firefly is trained on licensed and owned material and sold with an indemnity for enterprise use — a legal position, bought deliberately, that no GPU bill or token rate can express. For a company that cannot risk generated assets carrying an unresolved training-data claim, that indemnity is the entire product and the image quality is secondary. Cheaper models exist and produce comparable pictures; they do not come with anyone standing behind them.", "usageAssumption": "400 image generations a month from the included credit allowance, priced against what the same volume would cost on a commodity hosted image API.", "valueAdd": [ "proprietary-model", "compliance-regulatory", "integrations" ], "whatYouLose": "The indemnity, and generation inside Photoshop and Illustrator rather than beside them. For a freelancer neither may matter; for a brand's legal team the first one is the whole decision.", "diyPath": "Flux, SDXL or any hosted image API produce comparable output for less. None of them indemnify you, and the training-data question stays yours to answer.", "sources": [ { "kind": "pricing", "label": "Adobe Firefly plans", "url": "https://www.adobe.com/products/firefly/plans.html" }, { "kind": "rate-card", "label": "Replicate pricing", "url": "https://replicate.com/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "chatgpt-plus", "name": "ChatGPT Plus", "vendor": "OpenAI", "category": "assistant", "page": "https://isitjustawrapper.com/app/chatgpt-plus/", "productUrl": "https://chatgpt.com", "tier": "Plus", "monthlyUsd": 20, "billingNote": "Billed monthly. Higher Pro and Business tiers exist and are not covered by this entry. OpenAI blocks automated requests to its pricing pages, so this figure cannot be machine-checked and is confirmed by hand.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 40.5, "markupMultiple": 0.49, "verdict": "fair", "verdictSummary": "Like other first-party assistant subscriptions, a daily user's conversations are estimated to cost more through the published API than the subscription charges.", "verdictReasoning": "This is the product most people mean when they say they pay for AI, and it is not a wrapper around anything — it is the model vendor selling access to its own model. The flat fee sits below the estimated API cost of the assumed usage, which is what a subsidised consumer tier looks like. Bundled tools and image generation are not priced in here at all, so the real gap is wider than shown.", "usageAssumption": "A daily user with 200 conversations a month, each carrying roughly 10,000 tokens of accumulated context and returning 700 tokens, answered by a large frontier model.", "valueAdd": [ "proprietary-model", "scale-infra", "distribution" ], "whatYouLose": "The specific models and everything bundled around them. There is no cheaper route to the same product, only routes to different products.", "diyPath": "The API with your own client, which is cheaper for light use and more expensive at the volume assumed here. Open-weight models running locally are free of marginal cost and are not the same thing.", "sources": [ { "kind": "pricing", "label": "ChatGPT pricing", "url": "https://openai.com/chatgpt/pricing/" }, { "kind": "rate-card", "label": "OpenAI API pricing", "url": "https://openai.com/api/pricing/" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "claude-pro", "name": "Claude Pro", "vendor": "Anthropic", "category": "assistant", "page": "https://isitjustawrapper.com/app/claude-pro/", "productUrl": "https://claude.com", "tier": "Pro", "monthlyUsd": 20, "billingNote": "$20/month billed monthly; $17/month equivalent on an annual plan billed at $200 up front. Max tiers start at $100/month and are not covered by this entry. Usage limits apply and tighten at peak times.", "priceCheckedOn": "2026-08-13", "estimatedApiCostUsd": 42.75, "markupMultiple": 0.47, "verdict": "fair", "verdictSummary": "At a daily user's volume the subscription costs less than the same conversations would through the vendor's own API, so the plan is priced below its own inference bill.", "verdictReasoning": "There is no wrapper here to price: the subscription buys access to the model the vendor trains, from the vendor. The arithmetic is only interesting for what it says about flat pricing — a heavy user is subsidised by lighter ones, and the published API rate for the same conversations exceeds the fee. Read the multiple as a statement about the included allowance rather than about the company's economics, which the usage limits exist to manage.", "usageAssumption": "A daily user with 150 substantial conversations a month, each carrying roughly 15,000 tokens of accumulated context and returning 800 tokens, answered by a large frontier model.", "valueAdd": [ "proprietary-model", "scale-infra", "ux-craft" ], "whatYouLose": "The specific model, its context handling, and the ability to use it without managing keys or a client. Nothing about this is reproducible with someone else's API, because the model is the thing you are buying.", "diyPath": "Use the API directly with any client. At light usage that is cheaper than the subscription; at the volume assumed here it is estimated to cost roughly twice as much, and you take on rate-limit handling and a client of your own.", "sources": [ { "kind": "pricing", "label": "Claude pricing", "url": "https://claude.com/pricing" }, { "kind": "rate-card", "label": "Anthropic API pricing", "url": "https://claude.com/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-08-13" }, "lastReviewed": "2026-08-13" }, { "slug": "google-ai-pro", "name": "Google AI Pro", "vendor": "Google", "category": "assistant", "page": "https://isitjustawrapper.com/app/google-ai-pro/", "productUrl": "https://gemini.google.com", "tier": "AI Pro", "monthlyUsd": 19.99, "billingNote": "$19.99/month. The tier bundles cloud storage and features across Workspace, so a subscriber is buying more than the assistant this entry prices. Google has repackaged and renamed it more than once, which makes this row go stale faster than most. Multi-month free trials are commonly offered.", "priceCheckedOn": "2026-09-18", "estimatedApiCostUsd": 48, "markupMultiple": 0.42, "verdict": "fair", "verdictSummary": "A first-party assistant subscription priced below what the same conversations would cost through a published API, before counting the storage and product bundling attached to it.", "verdictReasoning": "Like the other assistant subscriptions here, there is no wrapper to price: the vendor trains the model and sells access to it. The estimate is also unfair to Google in a way worth naming — the tier bundles storage and features across several products that no token rate captures, so the real gap is wider than shown. What the arithmetic does illustrate is how heavily consumer AI tiers are subsidised at the top of the usage distribution.", "usageAssumption": "A daily user with 200 conversations a month, each carrying roughly 12,000 tokens of accumulated context and returning 800, answered by a large frontier model.", "valueAdd": [ "proprietary-model", "scale-infra", "distribution" ], "whatYouLose": "The specific models, and the integration into mail, documents and search that is the actual reason most subscribers stay. None of that is reproducible with an API key.", "diyPath": "The same models are available through the API, which is cheaper for light use and more expensive at the volume assumed here. Open-weight models run locally at no marginal cost and are a different product.", "sources": [ { "kind": "pricing", "label": "Google One AI plans", "url": "https://one.google.com/about/google-ai-plans/" }, { "kind": "rate-card", "label": "Google AI pricing", "url": "https://ai.google.dev/pricing" } ], "verification": { "status": "verified", "checkedBy": "@dovisvai", "checkedOn": "2026-09-18" }, "lastReviewed": "2026-09-18" } ] }