Free calculator and live comparison across OpenAI, Anthropic, Gemini, Mistral, DeepSeek, Qwen and Moonshot. Paste invoices, connect provider keys, and track your real spend — no card required to start.
Every rate below is the provider's own published price per million tokens — updated July 2026. No markup, no estimate.
Paste raw billing text from any provider's dashboard. We extract the model, tokens, and cost — itemized in seconds.
Add API keys for OpenRouter, OpenAI, Anthropic, and more — encrypted at rest. We only ever call each provider's own usage endpoint.
Live credit balances, token consumption, and monthly spend history — across every provider, on one dashboard.
Blended rate assumes a 3:1 input-to-output ratio, a common chat workload.
No subscription to try tokn.lat. Create an account when you want your data to stick around.
Snapshot, refreshed periodically from each provider's official pricing page — verified July 2026. Providers change rates often, so confirm on the live pricing page before committing a production budget.
Keys are encrypted at rest before storage and only ever used server-side to call that provider's own usage API. They're never exposed to the browser or sent anywhere else.
OpenRouter, OpenAI, Anthropic, Google Gemini, Mistral, DeepSeek, Qwen, and Moonshot — Western and Chinese providers on one dashboard.
Subscription invoices (like Claude Pro) don't include per-model usage — we'll show you the subscription cost and note that a usage export is needed for the full breakdown.
No. Parsing runs through a language model to extract structure, then we store only the extracted numbers — not the raw text — unless you explicitly choose to keep it.
25 invoice parses per day per account. The calculator and comparison table have no limit at all, even without an account.
Connect your provider keys and tokn.lat tracks live usage, credits, and monthly spend automatically.
Create free account →