Command Code Provider API

Command Code provides API endpoints compatible with OpenAI and Anthropic formats. You can utilize your Command Code plan and credits for the available models using any client that supports those standards.

Endpoints

Endpoint	Method	Format
`https://api.commandcode.ai/provider/v1/chat/completions`	POST	OpenAI Chat Completions
`https://api.commandcode.ai/provider/v1/messages`	POST	Anthropic Messages
`https://api.commandcode.ai/provider/v1/models`	GET	Models list

Quickstart

Get an API key

Create your API key from Studio.

Command Code API requires the Pro plan or higher. See Pricing & Limits for plan tiers.

Send a request

Once you have your API key, you can send a request to one of the following endpoints.

OpenAI Chat Completions /v1/chat/completions

curl https://api.commandcode.ai/provider/v1/chat/completions \
  -H "Authorization: Bearer $CMD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-v4-flash",
    "messages": [{"role": "user", "content": "Write a haiku about race conditions."}]
  }'

Anthropic Messages /v1/messages

curl https://api.commandcode.ai/provider/v1/messages \
  -H "Authorization: Bearer $CMD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Write a haiku about race conditions."}]
  }'

Models

All Command Code models are available via the API, see Pricing & Limits for the full model list and per-model pricing. You can fetch the live models list at runtime from /provider/v1/models.

List models

curl https://api.commandcode.ai/provider/v1/models

Streaming

Streaming is supported and can be received by setting stream: true.

Streaming

curl https://api.commandcode.ai/provider/v1/chat/completions \
  -H "Authorization: Bearer $CMD_API_KEY" \
  -H "Content-Type: application/json" \
  --no-buffer \
  -d '{
    "model": "deepseek/deepseek-v4-flash",
    "messages": [{"role": "user", "content": "Count to 5."}],
    "stream": true,
    "stream_options": {"include_usage": true}
  }'

Both endpoints emit token usage at the end of every stream — OpenAI clients see a final usage chunk, Anthropic clients see usage in the message_delta event. No opt-in required.

Errors

OpenAI clients see the standard error envelope:

{
  "error": {
    "message": "…",
    "type": "invalid_request_error" | "authentication_error" | "permission_error" | "rate_limit_error" | "server_error",
    "code": "…",
    "param": "…"
  }
}

Anthropic clients see:

{
  "type": "error",
  "error": {
    "type": "invalid_request_error" | "authentication_error" | "permission_error" | "rate_limit_error" | "api_error",
    "message": "…"
  }
}

Here are the common error cases:

Status	Code / Type	When
`400`	`unsupported_model`	Model isn't in the catalog (OpenAI shape only — `code` is on the OpenAI envelope; the Anthropic envelope has `type: invalid_request_error` instead).
`400`	`invalid_request_error`	Wrong endpoint for the model (Claude on `/chat/completions` or non-Anthropic on `/messages`), bad request body, or malformed JSON.
`401`	`authentication_error`	Missing or invalid auth — pass `Authorization: Bearer <token>` (any route) or `x-api-key: <token>` (Anthropic SDK on `/messages`).
`403`	`upgrade_required`	Upgrade to Pro plan or higher.
`429`	`rate_limit_error`	Rate-limited by the upstream. Retry with backoff.
`5xx`	`server_error` / `api_error`	Upstream failure. The body carries the upstream provider's error message.

FAQs

Which plans can use the Provider API?

Pro plan or higher. Accounts on the Go plan get a 403 upgrade_required on every /provider/v1/* request and need to upgrade before they can call the API.

Can I get multiple completions per request?

No. The endpoint returns one completion per request, n > 1 is rejected at the schema layer. Send parallel requests if you want multiple samples.

What content types are supported on user messages?

Text and image. Audio, file, and document parts are rejected by the schema. If the chosen model doesn't support vision the upstream provider returns its own error (we don't pre-gate per model).

Which models can I call?

Anthropic, OpenAI, and OSS models. See the Models section. Each endpoint accepts only its native wire shape: /chat/completions serves OpenAI and OSS models; /messages serves Anthropic models. Sending a Claude model to /chat/completions returns a 400 directing you to /messages, and vice versa.

Do I need a separate API key for this?

No. Use the same Command Code API key you generate for the CLI / Studio. Get one at CommandCode Studio under Settings → API Keys.

Reference

Command Code Provider API

Endpoints

Quickstart

Models

List models

Streaming

Streaming

Errors

FAQs