Use these endpoints when you need the native Bailian request format ({model, input, parameters}); for plain Qwen chat we recommend the OpenAI-compatible Chat Completions. Available models come from List Models (protocols contains dashscope).
Authorizations
Authorizationstring必填
Bearer Token authentication.
Authorization: Bearer YOUR_API_KEYEndpoints
POST /api/v1/services/aigc/text-generation/generation # text generationPOST /api/v1/services/aigc/multimodal-generation/generation # multimodal generationRequest headers
| Header | Required | Description |
|---|---|---|
Authorization |
Yes | Bearer YOUR_API_KEY |
Content-Type |
Yes | application/json |
X-DashScope-SSE |
Required for streaming | Set to enable to turn on SSE streaming (native Bailian semantics, not a body field); incremental output is controlled by parameters.incremental_output |
X-DashScope-DataInspection |
No | Enhanced content-safety inspection (official optional Bailian header) |
Notes
- Request/response fields follow the official Bailian docs; no field rewriting
- Service/action combinations outside the whitelist return 404
unsupported_endpoint - Async image / video: see DashScope Image Generation and DashScope Video Generation