中文文档

接口 Ask for one submitted request to be cancelled. Asks Comfy to stop a request that has not finished. It is a request, not a guarantee, and the `202` says exactly that: `CANCELLATION_REQUESTED` means the ask was accepted, not that the run has stopped… ask-for-one-submitted-request-to-be-cancelled 接口 Asset bytes Serves the bytes directly on surfaces where the platform stores blobs asset-bytes 接口 Asset metadata Returns the asset object with a fresh short-lived `url` for the content. Re-fetching always yields fresh URLs. asset-metadata 接口 Collect the result of one submitted request. The collect endpoint. On a request that has finished successfully it returns the partner model's own native output, byte for byte what the synchronous route's `200` carries for the same model and the … collect-the-result-of-one-submitted-request 接口 Create or reuse an asset over existing bytes Zero-byte fast-path: creates or reuses an asset over a blob the create-or-reuse-an-asset-over-existing-bytes 接口 Delete an asset record Deletes the asset RECORD. The underlying content-addressed blob is delete-an-asset-record 接口 Existence check for a blob by blake3 hash `200` if the calling account can mint from this blob, `404` otherwise. Same account-scoping as `from-hash`; lets a client decide between the dedup fast-path and a full upload before sending bytes. Use… existence-check-for-a-blob-by-blake3-hash 接口 Job status (the polling workhorse) Returns the full job object: current status, the latest progress job-status-the-polling-workhorse 接口 List jobs, newest first Lists jobs newest first. At a deployment's address, the jobs sent to list-jobs-newest-first 接口 List the models Comfy Router can run. Comfy Router's model catalog - one page of the canonical model IDs that `POST /v2/models/{provider}/{model}` accepts. An SDK calls this on cold start to discover what is runnable, and the `model_not_f… list-the-models-comfy-router-can-run 接口 Live event stream (SSE) Server-Sent Events stream of the job's live state. On connect the live-event-stream-sse 接口 Comfy API v2 概览 官方 Comfy API v2 参考：从外部应用上传输入、提交工作流任务并轮询获取结果，在 ComfyUI 中运行工作流。 overview 接口 Price a model call before running it, without running it. Returns a pre-flight estimate, in US dollars, of what running this body on this model will cost, priced at Comfy's standard list price (a model billed the cost its vendor reports is estimated from the… price-a-model-call-before-running-it-without-running-it 接口 Read one partner model's catalog entry by canonical model ID. Per-model detail for a single Comfy Router model, so a caller can check one model without walking the whole paginated catalog. The SDKs use it to look a model up immediately before invoking it. read-one-partner-models-catalog-entry-by-canonical-model-id 接口 Read one partner model's input and output schemas as an OpenAPI document. The per-model input and output schemas for a single Comfy Router model, served as a standalone OpenAPI document, so a caller - an SDK, a codegen tool, or an agent - can discover a model's arguments, a… read-one-partner-models-input-and-output-schemas-as-an-openapi-document 接口 Read the queue state of one submitted request. The poll endpoint. It answers with the request's current state and never with the result, so a client can watch a long generation without transferring its output on every poll - the result is collecte… read-the-queue-state-of-one-submitted-request 接口 Request cancellation Requests cancellation and returns the current job object — request-cancellation 接口 Run a partner model synchronously by canonical model ID. Comfy Router's canonical, model-ID-addressed entry point. The request body is the partner model's own native JSON input and the success response is that model's own native output — JSON for most model… run-a-partner-model-synchronously-by-canonical-model-id 接口 Submit a partner model run to the queue and return immediately. Comfy Router's queued delivery mode. The request body is the same partner-native JSON input `POST /v2/models/{provider}/{model}` accepts for this model - one body shape, one per-model schema, two deli… submit-a-partner-model-run-to-the-queue-and-return-immediately 接口 Submit a workflow for execution Accepts the API-format workflow graph verbatim. Validation is submit-a-workflow-for-execution 接口 Upload an asset (single-call multipart) Single-call `multipart/form-data` upload. The platform streams the upload-an-asset-single-call-multipart 接口 What the run printed Returns the job's captured execution log. Fetched on demand: a log what-the-run-printed