Process message batches

documented

Process large volumes of Messages requests asynchronously with 50% cost reduction.

POST /v1/messages/batchesMessagesno implementation yet0 runs

Input

custom_id*

Request custom_id

Unique per request within batch

model*

Model ID

The Claude model to use (e.g., claude-opus-4-6)

max_tokens*

Max tokens

Maximum tokens to generate

user_message*

User message content

Content of the user message

Calls Anthropic as you (your API key is your account); keychains.dev holds the credentials.