--- title: "Create an Anthropic message" method: POST path: "/messages" tags: ["Messages API"] --- # Create an Anthropic message `POST /messages` Anthropic-compatible Messages endpoint supporting system prompts, tool calling, and streaming (SSE). ## Headers - `anthropic-idempotency-key` string ## Request body - CreateMessageRequest - `model` string, required - `max_tokens` integer, required - `messages` AnthropicMessageParam[], required - `role` 'user' | 'assistant' | 'system' | 'developer', required — user/assistant per Anthropic; system and developer are also accepted (Claude Code injects mid-conversation system/developer messages). - `content` union, required - string - union[] - union - AnthropicTextBlock - `type` 'text', required - `text` string, required - `cache_control` AnthropicCacheControl — Prompt-cache hint (e.g. { "type": "ephemeral" }). Accepted on content blocks for Anthropic SDK compatibility but ignored — no cache read/write; see the Prompt caching note. - … - AnthropicImageBlock — Image content block. Image input is supported only on multimodal models; see the Models page. - `type` 'image', required - `source` union, required - … - `cache_control` AnthropicCacheControl — Prompt-cache hint (e.g. { "type": "ephemeral" }). Accepted on content blocks for Anthropic SDK compatibility but ignored — no cache read/write; see the Prompt caching note. - … - AnthropicToolUseBlock — Tool call emitted by the model. Return its outcome as a tool_result block in a follow-up user message. - `type` 'tool_use', required - `id` string, required - `name` string, required - `input` object, required - `cache_control` AnthropicCacheControl — Prompt-cache hint (e.g. { "type": "ephemeral" }). Accepted on content blocks for Anthropic SDK compatibility but ignored — no cache read/write; see the Prompt caching note. - … - AnthropicToolResultBlock — Result of a tool call, sent back in a user message. - `type` 'tool_result', required - `tool_use_id` string, required - `content` union, required - … - `is_error` boolean - `cache_control` AnthropicCacheControl — Prompt-cache hint (e.g. { "type": "ephemeral" }). Accepted on content blocks for Anthropic SDK compatibility but ignored — no cache read/write; see the Prompt caching note. - … - AnthropicThinkingBlock — Extended-thinking reasoning block. When continuing a thinking conversation, replay it unchanged (including its signature). - `type` 'thinking', required - `thinking` string, required - `signature` string — Opaque signature returned with the block; replay unchanged. - `cache_control` AnthropicCacheControl — Prompt-cache hint (e.g. { "type": "ephemeral" }). Accepted on content blocks for Anthropic SDK compatibility but ignored — no cache read/write; see the Prompt caching note. - … - AnthropicRedactedThinkingBlock — Redacted extended-thinking block. Accepted when replayed; the opaque encrypted reasoning is not used. - `type` 'redacted_thinking', required - `data` string, required — Opaque, encrypted reasoning payload. - `cache_control` AnthropicCacheControl — Prompt-cache hint (e.g. { "type": "ephemeral" }). Accepted on content blocks for Anthropic SDK compatibility but ignored — no cache read/write; see the Prompt caching note. - … - `system` union — System prompt: a string, or an array of text blocks. - string - AnthropicTextBlock[] - `type` 'text', required - `text` string, required - `cache_control` AnthropicCacheControl — Prompt-cache hint (e.g. { "type": "ephemeral" }). Accepted on content blocks for Anthropic SDK compatibility but ignored — no cache read/write; see the Prompt caching note. - `type` 'ephemeral' - `temperature` number - `top_p` number - `output_config` AnthropicOutputConfig - `format` AnthropicOutputConfigFormat - `type` 'json_schema', required - `schema` object, required - `name` string - `strict` boolean - `effort` 'none' | 'minimal' | 'low' | 'medium' | 'high' | 'xhigh' | 'max' — Reasoning effort for the model (`max` is treated as `xhigh`). Overrides any effort derived from `thinking`. - `tools` object[] — Tool definitions the model may call. Responses include `tool_use` blocks; return outcomes as `tool_result` content blocks in a follow-up message. - `tool_choice` object — Controls tool use: `auto`, `any`, `tool`, or `none`. - `thinking` object — Extended thinking configuration; translated to the model's reasoning. - `stream` boolean — When `true`, the response streams as Anthropic Server-Sent Events (`message_start`, `content_block_delta`, `message_stop`, …). - `metadata` RequestMetadata — Optional string metadata. completion_window controls scheduling; completion_webhook/webhook_token configure completion webhooks. - `completion_window` 'asap' | 'priority' | 'standard' | 'flex' — [Completion window](/completion-windows) (i.e. latency tier) for the request. Support matrix is available on the [Pricing](/pricing) page. When omitted, defaults to `standard` if that window is supported for the model; otherwise `flex` for async requests when the model supports it, and `asap` in all other cases. See [default behavior](/completion-windows#default-behavior). - `completion_webhook` string, uri - `webhook_token` string - `context_management` object — Anthropic server-side context-editing config (Claude Code sends it automatically). Accepted for compatibility but ignored — Sail forwards the full context each turn. ## Response `200` Anthropic-compatible message response. Returns a single JSON object by default, or an Anthropic Server-Sent Events stream when stream: true. - AnthropicMessageResponse - `id` string, required - `type` 'message', required - `role` 'assistant', required - `content` union[], required - union - AnthropicTextBlock - `type` 'text', required - `text` string, required - `cache_control` AnthropicCacheControl — Prompt-cache hint (e.g. { "type": "ephemeral" }). Accepted on content blocks for Anthropic SDK compatibility but ignored — no cache read/write; see the Prompt caching note. - `type` 'ephemeral' - AnthropicToolUseBlock — Tool call emitted by the model. Return its outcome as a tool_result block in a follow-up user message. - `type` 'tool_use', required - `id` string, required - `name` string, required - `input` object, required - `cache_control` AnthropicCacheControl — Prompt-cache hint (e.g. { "type": "ephemeral" }). Accepted on content blocks for Anthropic SDK compatibility but ignored — no cache read/write; see the Prompt caching note. - `type` 'ephemeral' - AnthropicThinkingBlock — Extended-thinking reasoning block. When continuing a thinking conversation, replay it unchanged (including its signature). - `type` 'thinking', required - `thinking` string, required - `signature` string — Opaque signature returned with the block; replay unchanged. - `cache_control` AnthropicCacheControl — Prompt-cache hint (e.g. { "type": "ephemeral" }). Accepted on content blocks for Anthropic SDK compatibility but ignored — no cache read/write; see the Prompt caching note. - `type` 'ephemeral' - `model` string, required - `stop_reason` 'end_turn' | 'max_tokens' | 'tool_use' | 'stop_sequence' | 'refusal', required - `stop_sequence` string, nullable, required - `usage` AnthropicUsage, required - `input_tokens` integer, required - `output_tokens` integer, required ## Other responses - `400` — Invalid request or unsupported feature. - `401` — Authentication error. - `408` — Timed out waiting for completion. - `500` — Server error. --- [API](https://skmtc.dev/sailresearch/apis/sail-api.md) · [All operations](https://skmtc.dev/sailresearch/apis/sail-api/llms.txt) · [OpenAPI document](https://skmtc-service-production.skmtc.workers.dev/v1/apis/sailresearch/sail-api/revisions/c2a783d507f9/schema)