canonical: https://jentic.com/apis/dashscope.aliyuncs.com/dashscope-aliyuncs-com

# Dashscope Aliyuncs Alibaba Cloud DashScope Qwen API (OpenAI-Compatible)

Jentic publishes the only available OpenAPI specification for Alibaba Cloud DashScope Qwen API (OpenAI-Compatible), keeping it validated and agent-ready. The service exposes Alibaba Cloud's Qwen (Tongyi Qianwen) models through an OpenAI-compatible interface for chat completions, text embeddings, and file management, authenticated with a bearer API key. It is available from mainland China, international, and US regional hosts.

## For AI agents

Lets an AI agent call Alibaba Cloud's Qwen models for chat completions and embeddings and manage uploaded files through an OpenAI-compatible endpoint.

## Scope

Does not fine-tune models, host inference endpoints, or manage Alibaba Cloud accounts: use it for Qwen chat completions, embeddings, and file management only.

## Capabilities

- Generate chat completions with Qwen models
- Create text embeddings for semantic search
- Upload files for processing or batch tasks
- List and retrieve uploaded file metadata
- Delete uploaded files

## Use cases

### Qwen chat integration

Products adding a chat assistant can reach Qwen through the OpenAI-compatible interface. The chat completions operation accepts the familiar request shape, so existing client code can target Qwen with a changed base URL and key.

Example prompt: Send a chat completion request to a Qwen model and return the assistant's reply

### Embeddings for retrieval

Semantic search and retrieval pipelines need vector embeddings. The embeddings operation turns text into vectors with Qwen embedding models, so a pipeline can index documents for similarity search.

Example prompt: Create embeddings for a batch of documents and return the vectors for indexing

### Agent-driven document workflow

An agent handling documents can upload inputs and reason over them. Through Jentic, an agent uploads a file, then runs chat or embeddings against its content, and removes the file when finished.

Example prompt: Upload a file, run a chat completion referencing it, then delete the file when done

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| POST | `/chat/completions` | Create a chat completion |
| POST | `/embeddings` | Create text embeddings |
| GET | `/files` | List uploaded files |
| POST | `/files` | Upload a file |
| GET | `/files/{file_id}` | Get file metadata |
| DELETE | `/files/{file_id}` | Delete a file |

## Key resources

- **Chat completions** — Generate chat completions with Qwen models
- **Embeddings** — Create text embeddings
- **Files** — Upload, list, retrieve, and delete files for processing or batch tasks

## AI readiness

This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.

- **Score:** 68 / 100
- **Maturity:** AI-Aware
- **Dimensions:**
  - Foundational Compliance: 93 / 100
  - Developer Experience & Jentic Compatibility: 63 / 100
  - AI-Readiness & Agent Experience: 49 / 100
  - Agent Usability: 94 / 100
  - Security: 60 / 100
  - AI Discoverability: 88 / 100
- **View full report:** https://jentic.com/apis/dashscope.aliyuncs.com/dashscope-aliyuncs-com/scorecard
- **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/
- **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview

### Score it yourself

Every API in the directory is allowlisted, so you can re-score it with no key required.

- **Score your own API:** https://jentic.com/scorecard.md
- **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md

```sh
npx @jentic/api-scorecard-cli score <openapi-url>
```

## Why Jentic

- **Setup:** DashScope follows the OpenAI-compatible shape but splits across three regional hosts and needs a bearer key. With Jentic One you install once, import the operations from the API Directory, and keep the key in a single place.
- **Permission scoping:** The operations are chat, embeddings, and file management, so you can scope an agent to inference only and keep file deletion out of its reach.
- **Credential handling:** Your DashScope API key is stored encrypted by your own Jentic One instance and injected when a call runs, so it never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents find the right Qwen operation through Jentic's intent search, matching a request like generating a chat completion to the matching DashScope endpoint.

## Related APIs

- **OpenAI** — The OpenAI API whose shape DashScope mirrors
- **Mistral** — Open-weight and hosted chat and embedding models
- **DeepSeek** — Chat and reasoning models with an OpenAI-style API

## FAQ

### How does the DashScope Qwen API authenticate?

Requests carry a DashScope API key as a bearer token, matching the OpenAI-compatible convention. Jentic One stores the key and injects it at execution time.

### Can I limit what my agent can do with the DashScope API?

Yes. You import only the operations an agent needs, so a chat agent can run completions and embeddings while file deletion stays out of scope.

### Which regional host should I use for DashScope?

DashScope is reachable from mainland China, international, and US hosts; pick the regional base URL that matches your account and data-residency needs.

### Is this really OpenAI-compatible?

Yes. The chat completions and embeddings operations follow the OpenAI request and response shapes, so existing OpenAI client code can target DashScope by changing the base URL and key.

### Why is there no official OpenAPI spec for DashScope's Qwen API?

Alibaba Cloud documents the OpenAI-compatible interface in prose rather than a machine-readable spec. Jentic generated this OpenAPI description from that documentation and keeps it validated so agents can call it reliably.