Create Message
AI-generatedSummary
Generate a message through CLI Proxy Web’s Anthropic-compatible POST /v1/messages endpoint using a model, conversation messages, and an optional system prompt.
Inputs
- messageModel — Model identifier sent in the request; defaults to claude-sonnet-4-5.
- messageMaxTokens — Maximum generation tokens sent as max_tokens; defaults to 1024, with a UI minimum of 1.
- messageSystem — Optional system prompt. Whitespace is trimmed, and blank values are omitted.
- messageMessages — Anthropic-compatible conversation messages. Must be a JSON array; defaults to a single user message saying 'Hello from n8n.'.
- messageOptions — Optional settings: Stream requests streaming from the proxy; Temperature (0–2) and Top P (0–1) are sent when selected. Additional Body must be a JSON object and is merged last, overriding any matching request fields.
Output shape
One n8n JSON item per input item containing the proxy response.
Response objects pass through unchanged; arrays are wrapped as {data: [...]}, and scalar or null responses as {value: ...}. Streaming is only a request flag: the node does not parse or emit individual token events. With Continue On Fail enabled, failures return an item containing an error string; otherwise execution throws.