Skip to content
AI Coding Tools & Models

Use Qwen3.8-Max With Codex CLI on WSL

Run Qwen3.8-Max in Codex CLI through Requesty's Responses-compatible endpoint on WSL, with the current model ID and config.

MGMCSA Guru Team August 18, 2026 2 min read
Codex CLI in WSL using Qwen3.8-Max through Requesty

The older version of this guide used Codex’s Chat Completions compatibility mode with qwen3-coder-plus. Current Codex is built around the Responses API. Requesty documents a Responses-compatible endpoint that translates requests for non-OpenAI providers, and its live catalog lists qwen3.8-max as the managed Qwen ID.

Create a Requesty key

Create a Requesty API key and confirm that qwen3.8-max is allowed for it. An access-list policy can hide or reject a model even when it appears in the public catalog.

Configure Codex

Open ~/.codex/config.toml in WSL and add:

model = "qwen3.8-max"
model_provider = "requesty"

[model_providers.requesty]
name = "Requesty"
base_url = "https://router.requesty.ai/v1"
env_key = "REQUESTY_API_KEY"
wire_api = "responses"

Set the key in the same shell that launches Codex:

export REQUESTY_API_KEY="your-requesty-key"
codex

Test the route

Start in a small repository and ask Codex to read a file without editing it. If that succeeds, request a small change and inspect the diff. Qwen3.8-Max supports tool calling, vision, reasoning, and long context, but the Codex client determines which of those features it sends through a custom provider.

Troubleshooting

  • A provider-policy error means the Requesty key or group does not allow qwen3.8-max.
  • A 401 response means REQUESTY_API_KEY is missing or invalid.
  • A protocol error usually means wire_api is not set to responses.
  • If requests use the wrong provider, check the top-level model_provider = "requesty" line.

Qwen and Codex checklist

  • Requesty key created
  • qwen3.8-max allowed for the key
  • Requesty base URL configured
  • wire_api set to responses
  • A read-only test completed

For a direct Alibaba connection without Codex, see Qwen3.8-Max with Claude Code or Qwen Code CLI.

Frequently asked questions

Can current Codex CLI call Qwen directly through DashScope?

Current Codex uses the Responses API, while Alibaba's standard OpenAI-compatible route is documented for chat compatibility. Requesty provides a Responses endpoint that converts requests for non-OpenAI models, which makes it the documented route used here.

Which Requesty model ID should I use?

Use qwen3.8-max. Requesty's live catalog lists it as the managed ID and can route it across available providers.

What context window is available?

Requesty and Alibaba list up to 1 million tokens of context and about 131,000 output tokens. A particular endpoint or client may enforce a smaller limit.

Sources & further reading

Official vendor documentation referenced while writing this guide.

Qwen3.8-Max Codex CLI Requesty WSL AI Coding
MG

MCSA Guru Team

IT & Systems Administration

We are working IT pros and system administrators who spend our days in Windows Server, Microsoft 365, and the wider Microsoft stack. MCSA Guru is where we write down the fixes and walkthroughs we wish we had found the first time.

About MCSA Guru

MCSA Guru provides independent, educational IT guidance. Microsoft, Windows, Windows Server, Microsoft 365, Exchange, and Microsoft Teams are trademarks of Microsoft Corporation; Docker is a trademark of Docker, Inc. MCSA Guru is not affiliated with or endorsed by Microsoft or Docker. Always test changes in a safe environment before applying them in production.

Related guides

Fixing something right now?

Jump straight into the guide library or search for the exact error or task you are dealing with.

Browse all guides Search the site