Cohere Free Trial API Key: Limits & Rate Limits (2026)
Free trial API key: 1,000 calls/month, all models, non-commercial
Cohere creates a free trial API key automatically when you sign up. Trial keys can use all of Cohere's models and APIs, but they are capped at 1,000 API calls per month and have per-endpoint rate limits, such as 20 requests per minute for Chat and 10 for Rerank. Cohere says trial keys may not be used for production or commercial purposes; for that you complete the Go to Production form and pay per use.
Who is eligible
- Cohere dashboard account (personal account; trial accounts can't add team members until they join an organization)
- Non-production, non-commercial use only
How to claim it
- Sign up at dashboard.cohere.com
- A trial key is created automatically; find it under API Keys
- Use it with the Cohere SDK or REST API, staying within 1,000 calls per month
Limits
| Monthly cap | 1,000 API calls/month |
|---|---|
| Chat (Command A, Command A+, Command A Reasoning, Command R, R+, R7B and others) | 20 requests/min |
| Embed (text) | 2,000 inputs/min |
| Embed (images) | 5 inputs/min |
| Rerank | 10 requests/min |
| EmbedJob | 5 requests/min |
| Audio transcriptions | 5 requests/min |
| Tokenize | 100 requests/min |
| Other endpoints (including Parse) | 500 requests/min |
Catches & caveats
Cohere's pricing page does not list per-token prices for Command A, Command R7B or Embed, so we don't quote them. The day the monthly trial cap resets is not published.
Cohere's free option is a trial API key that is created for you when you sign up. It works with all of Cohere's models and APIs (Command chat models, Embed, Rerank, audio transcription and more), but it is capped at 1,000 API calls per month, each endpoint has its own per-minute limit, and Cohere says trial keys may not be used for production or commercial purposes. It is an evaluation key, not a free production tier. When you are ready to ship, you complete the Go to Production form and pay per use.
What the trial key includes
Cohere offers two kinds of API keys. Its rate-limit documentation describes them as evaluation (trial) keys, which are free but limited in usage, and production keys, which are paid and much less limited. The trial key gives you:
- Access to all models and APIs. Cohere's pricing documentation says anyone can use all of Cohere's models and APIs with a trial key.
- 1,000 API calls per month across endpoints. This is a count of calls, not tokens, so a long chat request and a short one count the same.
- Per-minute rate limits by endpoint, listed in the next section.
- No charge. The pricing page lists the Trial plan as free, with the key created automatically at signup.
The trial is tied to a personal account. Cohere's pricing page notes that trial accounts are personal only and you cannot add members until you join an organization.
Trial vs production rate limits
These are the limits on Cohere's rate-limit page as of October 2026. Trial limits apply to every trial key; production limits apply once you have a paid production key.
| Endpoint or model | Trial key | Production key |
|---|---|---|
| Chat: Command A, Command R+, Command R, Command R7B, North Mini Code | 20 requests/min | 500 requests/min |
| Chat: Command A+, Command A Reasoning, Command A Translate, Command A Vision | 20 requests/min | Contact sales |
| Embed (text) | 2,000 inputs/min | 2,000 inputs/min |
| Embed (images) | 5 inputs/min | 400 inputs/min |
| EmbedJob | 5 requests/min | 50 requests/min |
| Rerank | 10 requests/min | 1,000 requests/min |
| Audio transcriptions | 5 requests/min | Contact sales |
| Tokenize | 100 requests/min | 2,000 requests/min |
| Parse | 500 requests/min | 500 requests/min |
| Everything else (default) | 500 requests/min | 500 requests/min |
The 1,000-calls-per-month cap is the limit most people hit first. At 20 chat requests per minute you could use the whole month's allowance in under an hour of continuous testing.
Cohere also notes that production keys on newer Chat model variants, such as Command A Reasoning, behave like trial keys, including the 1,000 calls a month limit, and asks you to contact sales@cohere.com before using those models in production. For a rate-limit increase on an existing production key, Cohere points to support@cohere.com.
What 1,000 calls a month is good for
The cap is enough to evaluate Cohere properly if you plan your tests. Some realistic budgets:
- Comparing chat models: 50 prompts × 4 Command models = 200 calls, leaving 800 for iteration.
- Testing Embed on a sample corpus: Embed calls accept batches of inputs, and the per-minute limit is counted in inputs (2,000/min). Embedding 10,000 short texts in batches of 90 takes about 112 calls, instead of 10,000 calls one by one.
- Evaluating Rerank: 300 queries against your retrieval results is 300 calls. At 10 per minute that takes about 30 minutes to run.
- A demo or hackathon prototype: fine for personal demos. A public app with real users would both breach the non-commercial rule and run out of calls quickly.
Batch your Embed requests. Because the monthly cap counts calls, sending many inputs per Embed call is the single biggest way to stretch the trial.
How to get and use the trial key
- Sign up at dashboard.cohere.com.
- Open API Keys in the dashboard. A trial key is created automatically; copy it.
- Store the key in an environment variable (for example `CO_API_KEY`) rather than in source code.
- Install the SDK (`pip install cohere`) and make a test call.
- Watch your usage in the dashboard so you know how many of your 1,000 monthly calls remain.
import cohere
co = cohere.ClientV2() # reads CO_API_KEY
resp = co.chat(
model="command-a-03-2025",
messages=[{"role": "user", "content": "Summarize our refund policy in one line."}],
)
print(resp.message.content[0].text)Use the model name shown in Cohere's model documentation for the model you want to test; names change as new versions ship.
Moving to production
When you outgrow the trial, or want to use Cohere in a product, you need a production key. Cohere's Going Live documentation describes the process:
- An Owner of your Cohere team completes the Go to Production form in the dashboard. Only Owners can do this.
- You acknowledge Cohere's SaaS agreement and terms of service, and your organization confirms it has read the model limitations, model cards and data statement.
- You answer whether your use involves any of the sensitive use cases in Cohere's usage guidelines.
- If not, you can create and use production keys immediately. If you do declare a sensitive use case, the production key may be rate limited like a trial key until Cohere's safety team contacts you; Cohere says those reviews take no longer than 72 business hours.
Production usage is billed pay-as-you-go. Cohere's pricing page says bills are issued at the end of each calendar month or when the outstanding balance reaches $250. Only team Owners can view invoices and update payment information.
What paid usage costs
Cohere's pricing documentation says Command models are priced per token, with different rates for input and output tokens; Embed is priced on the number of tokens embedded; and Rerank is priced per search. A search is one query with up to 100 documents, and documents over 500 tokens (including the query) are split into chunks that each count as a document.
The public pricing page lists per-token prices only for some legacy models, for example:
| Model (legacy) | Input / 1M tokens | Output / 1M tokens |
|---|---|---|
| Command R 03-2024 | $0.50 | $1.50 |
| Command R+ 08-2024 | $2.50 | $10.00 |
| Command R+ 04-2024 | $3.00 | $15.00 |
| Aya Expanse 8B and 32B | $0.50 | $1.50 |
Current models such as Command A, Command R7B and Embed are not priced on that page, so check the dashboard or ask Cohere sales before estimating production costs. Dedicated Model Vault instances are priced per hour or per month; for example Embed 5 Small is listed at $3.00/hour or $2,000/month, and Rerank 3.5 Medium at $5.00/hour or $3,250/month.
Trial key rules to take seriously
- No production or commercial use. This is stated on the pricing page. Putting a trial key behind a paid product, or behind a free product run by a business, breaks the terms even if you stay under 1,000 calls.
- Keys belong to a team. Production keys can only be created after an Owner completes the Go to Production form; regular Users can only view, create, copy and rename trial keys.
- Existing accounts can't simply be invited. Cohere's team docs say users who already have a Cohere account cannot be invited to a team through the dashboard and need to contact support to migrate.
- Newer models may stay capped. Even on a production key, newer Chat variants like Command A Reasoning keep the 1,000 calls a month limit until you arrange access with sales.
Alternatives if you need more free usage
If 1,000 calls a month is too tight, or you need something you can use commercially for free, these are the closest options on our site:
| Need | Option | Why |
|---|---|---|
| Free embeddings at volume | Voyage AI free tokens | 200M free tokens per current model and on rerank-3 |
| Free embeddings plus reranker | Jina AI free API tokens | Free tokens usable across Jina models |
| Free chat model API | Gemini API free tier | Rate-limited free access to several Gemini models |
| Free chat with open models | Groq free tier | Rate-limited free access to open-weight models |
| Free European models | Mistral free API mode | Included monthly usage, no card |
| Many free models through one key | OpenRouter free models | Free model variants behind a single API |
Our free AI API keys guide compares these side by side.
Common questions from developers
Does each call count once, regardless of size?
Cohere states the cap as 1,000 API calls a month. It doesn't describe any token-based weighting for the trial cap, so a call counts as a call. The per-minute Embed limit, however, is counted in inputs, not calls.
Will I be charged if I exceed the trial cap?
Cohere says trial API calls are free and bills only production usage. Once you reach the cap, you either wait for the monthly allowance to come back or move to a production key.
Can I create several trial keys to get more calls?
You can create more than one trial key on the API Keys page, but Cohere describes the 1,000-call limit as applying to trial keys; it does not say the limit multiplies with the number of keys, so don't plan around that.
Where do I check Cohere's status if calls fail?
Cohere's Going Live docs point to its status page at status.cohere.ai, which lists incidents, component status and scheduled maintenance, and lets you subscribe to updates by email or phone. Check it before assuming you have hit a trial limit, since rate-limit errors and outages look similar from client code.
Frequently asked questions
›Is the Cohere API free?
Yes, for evaluation. Every account gets a free trial API key, capped at 1,000 API calls per month with per-endpoint rate limits. Production use requires a paid production key.
›Can I use a Cohere trial key commercially?
No. Cohere's pricing page says trial keys are not permitted to be used for production or commercial purposes.
›What are Cohere trial key rate limits?
Chat is 20 requests/min, Embed 2,000 inputs/min (5/min for images), Rerank 10 requests/min, EmbedJob 5/min, audio transcription 5/min and Tokenize 100/min, with 1,000 calls per month in total.
›Does the Cohere trial key give access to all models?
Cohere's docs say anyone can use all of Cohere's models and APIs with a trial key, within the trial limits.
›Do I need a credit card for a Cohere trial key?
Cohere's pricing page says the trial key is created automatically at signup and doesn't mention a card. Payment details are needed only for paid production usage.
›How do I upgrade from a trial key to a production key?
An account Owner completes the Go to Production form in the dashboard, accepting Cohere's SaaS agreement and terms and answering a question about sensitive use cases. Production keys can then be created and are billed pay-as-you-go.
›Does a production key remove the 1,000-call cap?
For most models, yes. But Cohere's rate-limit page says production keys on newer Chat variants, such as Command A Reasoning, are limited like trial keys, and asks you to contact sales before using them in production.
›When does the monthly trial cap reset?
Cohere describes the cap as 1,000 API calls a month but does not say on which day the count resets.
Sources
Checked against these pages on Oct 17, 2026. Providers change offers often: always confirm on the official page before you rely on it.
- Official: Cohere rate limits(docs.cohere.com)
- Official: Cohere pricing(cohere.com)
- Official: Cohere docs, How does Cohere's pricing work?(docs.cohere.com)
- Official: Cohere docs, Going live(docs.cohere.com)
- Official: Cohere docs, Teams and roles(docs.cohere.com)
Similar deals
Google Cloud Free Trial
$300 Welcome credit valid for 90 days
New Google Cloud customers get a $300 Welcome credit to spend over a 90-day Free Trial on most Google Cloud services. You must add a valid payment method, but you are not billed during the trial. The credit cannot pay for the Gemini API in AI Studio or for partner models offered as managed APIs, and trial accounts cannot add GPUs, use Marketplace or request quota increases. If you don't upgrade, resources are stopped at the end and deleted after a 30-day grace period. Separately, the always-free Free Tier keeps running with no end date.
Deepgram Speech API
$200 free credit for new accounts, no credit card
Every new Deepgram account gets $200 in free credit on the Pay As You Go plan, with no credit card required. Deepgram says this is about 43,000 minutes (over 700 hours) of Nova transcription, and the credit works across speech-to-text, text-to-speech, the Voice Agent API and audio intelligence. Deepgram's pricing page says the credit lasts until used, but its developer docs say free promotional credits expire one year from signup.
AssemblyAI Speech-to-Text API
$50 free credits, no credit card
AssemblyAI gives new accounts $50 in free credits with no credit card required. AssemblyAI's pricing FAQ says this covers up to 185 hours of pre-recorded or 333 hours of streaming transcription, depending on the model. Free accounts can use pre-recorded transcription, Speech Understanding and streaming, but not the LLM Gateway, and have lower concurrency limits until they upgrade.
Gemini API / Google AI Studio
Free tier: Gemini 3.8 Flash, 3.6 Flash, 3.5 Flash-Lite, 2.5 Pro, 2.5 Flash, Gemma 4 and more, rate-limited
Google's Gemini API has a Free tier where many models are listed as 'Free of charge' for input and output, including Gemini 3.8 Flash, Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, Gemini 2.5 Pro, Gemini 2.5 Flash, Gemini Embedding 2 and Gemma 4. Newer flagship, image, video and music models (Gemini 4 Argon, Gemini 3.1 Pro Preview, Nano Banana, Veo 3.1, Lyria) are paid-only. Free-tier prompts and responses may be used to improve Google's products and can be read by human reviewers. Rate limits apply per project and are shown in Google AI Studio rather than published as a fixed table.