Feature Description

Please add Qwen/Qwen3.8-Flash to the /provider/v1/responses endpoint, or let us know whether that is planned.

Today /models reports it as /chat/completions only, while most Qwen 3.7 / 3.8 siblings expose both endpoints.

Use Case

Clients that speak the Responses wire format (e.g. Codex with wire_api = "responses") cannot use Qwen/Qwen3.8-Flash through the Provider API. Users who want this model with native Responses currently have to pick a different model or fall back to a Chat Completions client.

Additional Context

As of 2026-10-03:

• Qwen/Qwen3.8-Flash → /chat/completions only
• Qwen/Qwen3.8-Omni-Flash, Qwen/Qwen3.8-Max, Qwen/Qwen3.8-27B, Qwen/Qwen3.7-Flash → also /responses

Since same-generation siblings already support /responses, is this planned for Qwen/Qwen3.8-Flash? If it's upstream-gated, is there an ETA?

How important is this to you?

Nice to have