Paste your key. CapMap reveals every model, capability, and permission it unlocks. Instantly. Securely. Free.
Provide your token securely. Memory processing ensures nothing is logged or kept.
Our engine actively tests standard endpoints to definitively prove capabilities.
Review allowed models, rate limits, and actual permission denials transparently.
Instantly confirm your key is active and practically usable for generation tasks.
See every precise model variation your specific API tier genuinely unlocks.
Text, embeddings, image, audio — all verified with micro live endpoint tests.
We trigger real HTTP hits, not heuristic guesswork, for 100% accuracy.
Pull and decode hidden API token tracking straight from live response headers.
Calculate realistic spend metrics depending on your tested model allowances.
Yes. Keys are strictly processed in volatile runtime memory and immediately discarded after testing. We write absolutely nothing to persistent databases or server logs.
For providers that support open CORS (like Gemini), we bypass our backend entirely. The network fetch happens straight from your client window directly to the provider's endpoint.
We currently natively decode universally recognized keys from OpenAI, Google Gemini, Anthropic, Groq, and Perplexity by examining their distinct prefix signatures.
The core capability discovery engine is fully open source. You can view, audit, or host the underlying tests yourself by visiting our GitHub repository.
Zero logging. As developers ourselves, we structured the architecture to inherently prevent caching, analytics tracking, or environment variable leaks.
No more guessing which endpoints are enabled on your tier.
Analyze Your Key Free →Reach out directly — I read every message.
Usually responds within 24 hours.