Qwen Code via Tokonomix

//What is Qwen Code?

Qwen Code is an open-source agentic command-line coding tool (adapted from Gemini CLI) by the Qwen team. It reads your repo, edits files, runs commands and iterates — and although it ships tuned for Qwen models, it speaks the OpenAI protocol, so any OpenAI-compatible endpoint works.

• A terminal coding agent: explore, edit, run and iterate over your project.
• Declares models in ~/.qwen/settings.json under modelProviders, keyed by API protocol (openai, anthropic, …).
• Reads API keys from an envKey reference (or .env / shell export, which take priority).
• Open source and model-agnostic via the openai protocol.

//Why route Qwen Code through Tokonomix?

Qwen Code already speaks the OpenAI protocol, so Tokonomix slots in as one more openai provider — except this one unlocks 130+ models and a consensus option, not just one vendor.

• One openai provider entry, 130+ models — switch by changing model.name.
• Use tokonomix-consensus for the hard reasoning turns, a fast slug for routine edits.
• Per-call spend caps keep an agentic CLI within budget.
• Pin by hosting_region / origin_country for regulated code.

//Connect Tokonomix in settings.json

Add a Tokonomix entry under modelProviders.openai with the Tokonomix baseUrl and an envKey, then point model.name at the slug you want.

{
  "modelProviders": {
    "openai": [
      {
        "id": "tokonomix-consensus",
        "name": "Tokonomix Consensus",
        "baseUrl": "https://tokonomix.ai/api/v1",
        "envKey": "TOKONOMIX_API_KEY"
      }
    ]
  },
  "env": { "TOKONOMIX_API_KEY": "tok_live_your-key" },
  "security": { "auth": { "selectedType": "openai" } },
  "model": { "name": "tokonomix-consensus" }
}

//Verify it works

• Save ~/.qwen/settings.json and start Qwen Code in a scratch repo.
• Run a small edit and confirm the model used is your Tokonomix slug.
• Check the call and token usage in your Tokonomix dashboard.