Anthropic Messages API | 51cowork API Docs

Quick answer

Use x-api-key and anthropic-version; model must come from /v1/models, and count_tokens can estimate input before generation.

Generate: POST /v1/messages
Count: POST /v1/messages/count_tokens
Header: x-api-key
Stream: stream: true

Generation request

Messagesbash

curl "https://api.51cowork.com/v1/messages" \
  -H "x-api-key: <key-from-dashboard>" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "<model-from-v1-models>",
    "max_tokens": 256,
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Field	Rule
model	ID returned by /v1/models
max_tokens	Positive integer within model support
messages	At least one valid message
stream	Optional event stream

Estimate input tokens

count_tokens estimates input and does not generate a response. Use the same model and messages as the generation request.

Count tokensbash

curl "https://api.51cowork.com/v1/messages/count_tokens" \
  -H "x-api-key: <key-from-dashboard>" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{"model":"<model-from-v1-models>","messages":[{"role":"user","content":"Hello"}]}'

Compatibility boundary

Do not assume every model supports identical content, tools, or context fields.
When upstream rejects a field, remove the smallest unsupported option and retry only when safe.
Consume streaming events incrementally instead of buffering the full body.