Tool review · By Adam Johnstone · Updated 2026-09-13
TypingMind is the only tool here that is not a subscription. You pay once for the front end, bring your own provider API keys, and pay OpenAI or Anthropic directly for what you use. Whether that is cheaper depends entirely on how much you use it.
Plans and prices
Checked against official sources on 2026-09-13. Where the vendor does not publish a figure, this table says so rather than filling the gap.
| Plan | Price | Notes |
|---|---|---|
| Free | $0 | Chat history is not saved and a popup appears every few minutes. The lack of saved history is the real constraint, not a message cap. |
| Standard | $39 one-time | Removes ads and popups. Basic chat, AI agents, chat sharing, plugins, voice input. |
| Extended | $79 one-time | Adds web search, vision, image generation, text-to-speech and document upload. |
| Premium | $99 one-time | Against a $198 list price. Adds multi-model parallel chat, projects and folders, artifacts and canvas, unlimited plugins and MCP, free updates. |
| Bulk | $395 one-time | Shared key for up to 10 users across 50 devices at the Premium feature level. |
| Teams (cloud) | $83–$249/mo | Starter $83 for 5 seats plus $8 per extra seat; Growth $166; Professional $249 adds analytics, audit logs, RBAC and SSO. Business from $18,700/year. |
Model access
Whatever your API keys reach. Documented providers: OpenAI, Anthropic, Google Gemini, OpenRouter, DeepInfra, DeepSeek, Groq, Mistral, Moonshot, Perplexity, xAI and Z.ai, plus custom endpoints for Azure OpenAI and local models through LMStudio.
This is the key structural difference on this site. Every other tool includes model access in the price; TypingMind requires you to bring your own keys and states plainly that “price does not include API costs”. The upside is no cap of any kind — TypingMind imposes no message limits, only your provider does. The downside is an unpredictable bill that scales with context length rather than message count, which catches people out when they use multi-model comparison.
Limits
| Limit | Value |
|---|---|
| Message and query caps | None imposed by TypingMind — your provider's limits apply |
| Devices per licence key | 5, with more purchasable |
| Local chat storage | Bounded by device storage; “thousands of conversations” per the FAQ |
| Free cloud sync storage | 50MB, roughly 5,000 chats; 1GB and 5GB upgrades exist but are not priced publicly |
| Context window | Set by the model you choose, not by TypingMind |
| Money-back window | 14 days |
When a conversation fills the model's context, TypingMind summarises the older messages and keeps the summary rather than truncating silently — a toggle under Settings, Internal Prompts. One independent review models the economics at roughly $1 a month for light use on economical models, $12 for regular balanced use, and $110 for heavy use of premium models. On those figures a $99 licence beats a single $20 subscription if your API spend stays under about $12 a month.
Fit
Open questions
Every comparison has gaps. Listing them is more useful than guessing, and these are the things worth checking yourself during a trial.
Platform and data
Web app first. The desktop and mobile apps are progressive web apps rather than native builds — TypingMind's own FAQ confirms there are no native apps. Self-hosting is supported through a static package and a Docker deployment for teams. The changelog is the most actively maintained of any tool here, with entries through 11 September 2026.
The only privacy claim on this site that is structural rather than promissory. Chats, prompts and API keys are stored in your browser's local storage and IndexedDB, and API keys are never transmitted to TypingMind's servers — requests go directly from your device to the provider. Cloud sync is opt-in and clearly scoped. Conversations are “never used for training purposes”, and the documentation correctly devolves the real question to your provider, noting OpenAI does not train on API data, Anthropic does not without explicit agreement, and Google's unpaid tiers may. GDPR statement, DPA and EULA are published.
Sources
All figures on this page were checked on 2026-09-13. Prices and limits in this category change frequently — verify on the official page before paying. TypingMind licence pricing · TypingMind Teams pricing · TypingMind privacy policy · TypingMind training policy · TypingMind FAQ · TypingMind changelog
FAQ
For individuals it is a one-time purchase, not a subscription: Standard $39, Extended $79, Premium $99 against a $198 list price, and a $395 bulk licence for up to 10 users. TypingMind Teams is separate and recurring, from $83 a month for 5 seats up to $249, with a Business tier from $18,700 a year. None of these include model costs.
No. You bring your own API keys and pay OpenAI, Anthropic, Google or another provider directly for usage. TypingMind's pricing page states plainly that price does not include API costs. This is the main structural difference between TypingMind and every other tool compared on this site.
It depends on volume. One independent 2026 review models light use at about $1 a month in API costs, regular use at $12, and heavy premium-model use at $110. Against a $99 one-time licence, TypingMind beats a single $20-a-month subscription if your API spend stays under roughly $12 a month, and beats three such subscriptions under about $52.
In your own browser, using local storage and IndexedDB. API keys are never sent to TypingMind's servers and requests go directly from your device to the AI provider. Cloud sync is available but opt-in, and only then does TypingMind hold your chats. This is verifiable from the product's architecture rather than being only a policy promise.
Yes. There is a static self-host package and a Docker-based deployment for TypingMind Teams. The self-hosted team licence price is not published — it routes to a quote form.
Compare