qwen3.7-max Chat Completions
Applicable to the publicly available chat model in the Ali channel’s final whitelist: qwen3.7-max.
Supported fields: model, messages, stream, temperature, top_p, max_completion_tokens, stop, seed, tools, tool_choice, enable_search, response_format.
Authorizations
Pass Authorization: Bearer <token> in the request header.
Body
Model name. Only publicly available chat models in the current Ali channel's final whitelist are supported.
qwen3.7-max "qwen3.7-max"
List of conversation messages. Supports text content.
Whether to use streaming output. Defaults to false.
Sampling temperature, controls output randomness. Range [0, 2].
0 <= x <= 2Nucleus sampling probability threshold.
0 <= x <= 1Maximum length of model output, including thinking process and model response. When the model output exceeds this value, generation will stop early and the returned finish_reason will be length. Default and maximum values are the model's maximum output length. Unlike max_tokens which only limits the response part, max_completion_tokens limits the complete model output (thinking + response). Recommended for thinking-enabled models.
List of stop sequences.
Random seed for reproducible generation results.
List of tools for Function Calling.
Tool selection strategy. Possible values: auto, none, required.
Whether to enable search-augmented generation. Only supported by some models.
Response format, used to force the model to output JSON.