cohere
chat
embedding
Test your Cohere API key.
Verify a Cohere API key, list Command-R / Embed / Rerank models the key can call, and benchmark the round trip.
Stateless proxy — keys never logged, stored, or persisted. What happens to your key →
Detected
Cohere
What this key does
A Cohere key authenticates the v1 Cohere API — Chat (Command-R / Command-R+), Embed, Rerank, and the classify and summarize endpoints. Auth is Bearer; the Chat shape is Cohere-specific (not OpenAI-compatible).
How to get a Cohere API key
- Sign in at dashboard.cohere.com.
- Open API Keys and create a new trial or production key.
- Trial keys are free but rate-limited; production keys require billing.
- Paste the key here to confirm it works.
Common errors and fixes
- 401 Unauthorized: Key is invalid, revoked, or pasted with extra whitespace. Generate a new key from the provider console and try again.
- 403 Forbidden: Key is valid but lacks permission for this resource. Check project / org / workspace scope, or that billing is set up for this key.
- 429 Too Many Requests: You hit the per-minute or per-day rate limit. Wait a moment and retry, or upgrade your tier.
- 404 Not Found: The endpoint or model id changed. Check the provider docs for the current path and model identifier.
- 5xx: The provider is having issues. Check their status page before assuming the bug is yours.
Security best practices
- Store keys in an env var or secret manager — never commit them to a repo, even a private one.
- Restrict scope: prefer per-project or per-deployment keys over a single root key shared across services.
- Rotate on a schedule (90 days is a sane default) and immediately on suspected leak.
- Audit usage in the provider console after rotation to confirm the old key has zero traffic.
- Set per-key spend limits where the provider supports them, so a leaked key has a bounded blast radius.
Pricing at a glance
Command-R is priced for high-throughput RAG. Embed is among the cheapest large-context embedding models. Rerank is billed per search query, not per token.
FAQ
- What's the difference between trial and production keys?
- Trial keys are free but throttled and labeled as for evaluation only. Production keys hit standard pay-as-you-go pricing.
- Is Cohere OpenAI-compatible?
- No. The Chat shape is different (uses message + chat_history, not messages[]).
- Which Cohere endpoint is cheapest to call for validation?
- /v1/models. We use it for the latency benchmark and key validation here.
- Does Cohere train on my data?
- Production-tier requests are not used for training by default. Trial keys are subject to the trial data policy — read it before sending sensitive prompts.
- What's the cheapest Cohere chat model?
- command-r is the price/performance pick. command-r-plus is the heavy-reasoning tier.
- Can I run Cohere models on AWS?
- Yes — several are mirrored on Bedrock with AWS-side billing.
Test other providers
Related reading
- API key security best practices for LLMsHow to store, scope, rotate, and revoke LLM API keys without leaking them through git, logs, or shared environments.
- OpenAI vs Anthropic pricing in 2026A side-by-side breakdown of OpenAI and Anthropic per-token pricing, batch discounts, and prompt-caching savings.