Spomory
← All platforms

Qwen (阿里云百炼)

Alibaba Cloud's model API — code-level MCP tool, SSE, up to 10 servers.

阿里云百炼 (Alibaba Cloud Model Studio) is Qwen's developer API platform — this is for building your own app on top of Qwen's models, not a setting inside the consumer Tongyi Qianwen chat app. MCP tools are added at the API/code level, through the Responses API's tools array, not a GUI. 百炼 currently only speaks the SSE transport for MCP, which is why Spomory runs a dedicated SSE mount alongside its original Streamable HTTP endpoint. Verified end-to-end against the real production endpoint.

Code-level (API)

In your call to the Responses API, add an MCP tool entry pointing at Spomory's SSE mount:

{
  "tools": [
    {
      "type": "mcp",
      "server_protocol": "sse",
      "server_label": "spomory",
      "server_url": "https://api.yliuai.com/mcp-apikey-sse/",
      "headers": {
        "x-api-key": "<your key>"
      }
    }
  ]
}

百炼 caps a single request at 10 MCP servers total, and supports SSE only (not the newer Streamable HTTP) — this is exactly why the URL above ends in -sse instead of the plain /mcp-apikey/ path used by every other client on this site.

Verified with a real curl-based SSE handshake against production (initialize, then a tool call, both over the -sse mount) — confirmed hitting the same server as every other client on this page.

Where does <your key> come from?

Get one instantly with curl — fine for your own testing:

curl -X POST https://api.yliuai.com/users/register \
  -H "Content-Type: application/json" \
  -d '{"email": "you@example.com"}'
# the api_key in the response (starts with mck_) is shown only this once -- save it

That hands you a key with no proof you control the email — fine for testing, but for real use, get one through the email-confirmation form instead, which actually verifies the address.

Get a verified key on the Cloud & Apps page →

阿里云百炼's MCP docs →

Make it actually get used

Connecting the tools doesn't mean your client reaches for them on its own — it needs to be told to call search_memory before answering and add_memory after you share something worth keeping. SKILL.md is a short instructions file that does that. How you load it is specific to each client: some auto-discover a skills folder the same way Claude Code does, others only have a custom-instructions or system-prompt field instead. Verified for this one:

百炼 is a raw API platform with no file-based persistent instructions — include SKILL.md's content as a system-role message in every API call instead.

SKILL.md on GitHub →