unbiased ai

Qwen Code, with Pareto at the other end

Qwen Code talks to OpenAI-compatible endpoints, and Unbiased serves one. Export a base URL, a key and the model name pareto, start qwen as usual, and the same terminal agent reads, edits and runs your code. Behind the model name, Pareto selects frontier and open models for each request and blends their outputs into one answer.

Choose Personal · $10/month See the setup →

Tested live with Qwen Code 0.25.350M tokens a week

A demonstration of Qwen Code on Pareto, recreated from a real run. Not a screenshot.

Set it up in three variables

Qwen Code reads OpenAI-compatible settings from the environment, from a .env file or from ~/.qwen/settings.json. Start with the shell; move to the settings file once you want it every time.

  1. Get an Unbiased API key

    Choose a plan, then create a key in the Unbiased platform. New accounts are reviewed and approved by a person before the key works.

  2. Install Qwen Code

    Install it with npm install -g @qwen-code/qwen-code@latest or brew install qwen-code. We tested version 0.25.3; check yours with qwen --version.

  3. Export three variables

    Set OPENAI_BASE_URL to https://api.unbiased.ai/v1, with the /v1. Put your key in OPENAI_API_KEY and set OPENAI_MODEL to pareto.

  4. Start it with the OpenAI auth type

    Run qwen --auth-type openai. The flag picks the OpenAI-compatible provider over anything saved in settings, and it works for one-shot runs too: qwen --auth-type openai --yolo "fix the failing test".

  5. Keep it, if you like it

    Declare pareto under modelProviders in ~/.qwen/settings.json and keep the key in ~/.qwen/.env, as shown. Plain qwen then starts on Pareto.

Tested on October 20, 2026Qwen Code 0.25.3 on Pareto (pareto). It found the off-by-one in a failing Python test, edited the file and reran the test to green. Run record.
Shell
export OPENAI_BASE_URL=https://api.unbiased.ai/v1
export OPENAI_API_KEY=your-unbiased-api-key
export OPENAI_MODEL=pareto
qwen --auth-type openai
Or ~/.qwen/settings.json
{
  "modelProviders": {
    "openai": [
      {
        "id": "pareto",
        "baseUrl": "https://api.unbiased.ai/v1",
        "envKey": "UNBIASED_API_KEY"
      }
    ]
  },
  "security": { "auth": { "selectedType": "openai" } },
  "model": { "name": "pareto" }
}
And ~/.qwen/.env
# Qwen Code reads only the first .env it finds
UNBIASED_API_KEY=your-unbiased-api-key

Qwen Code loads one .env file, the first it finds: .qwen/.env or .env from the current folder upward, then ~/.qwen/.env, then ~/.env. A project .env will hide the one in ~/.qwen.

Qwen Code stays. The model changes.

Three settings change. The agent, its tools and its approval modes stay where you left them.

One model string, several models behind it

Pareto is Unbiased’s blended AI model. On each request it selects frontier and open models and blends their outputs into one answer, so Qwen Code talks to one model string, pareto, and never has to pick.

“When the frontier improves, Pareto improves; when frontier prices fall, you should benefit.” From the pricing page.

How a Pareto answer gets made →
An illustration, not a wiring diagram. The blend includes frontier and open models, and its composition can change.

What an agent task costs on Pareto

Agents make many model calls per task, so the price of a finished task matters more than the price of a token. The model card publishes both for Pareto on two public agent benchmarks.

Terminal-Bench 4.0Tasks an agent completes in a terminal
$0.48/task
50.8% solved
DeepSWE v1.1Changes an agent makes in a real codebase
$0.24/task
69.9% solved

Pareto 26.10, preliminary October 1, 2026 runs that may change; denominators and the cost method still require confirmation. Cost is the reported mean USD per task at API rates, not an estimate for your workload. Model card and method.

Pay for Qwen Code by the month

One API key works across Qwen Code and any other tool you point at Pareto. Pick a monthly allowance, or pay per token on Team.

Personal

$10/month

  • 50M tokens a week
  • 20 requests per minute
  • For one person
Choose Personal

Personal Max

$100/month

  • 600M tokens a week
  • 60 requests per minute
  • For one person
Choose Personal Max

Team

Pay as you go

  • Prepaid credits at published rates
  • 300 requests per minute
  • No weekly token cap
Get started with Team

Subscriptions are for one person and are not unlimited. New API accounts are reviewed by a person before access turns on, and commercial or production use needs a separate written agreement. Rates and limits are on Pricing.

Qwen Code on Pareto, questions answered

Qwen OAuth is gone. Do I need a QwenCloud or ModelStudio account?

No. The free Qwen OAuth tier ended on April 15, 2026 (see the Qwen Code authentication guide), and Qwen Code now asks you to pick a provider. With these settings it sends every request to Unbiased with your Unbiased key, and your Unbiased plan pays for them. Qwen Code itself is free and open source.

Which auth type should I choose?

Use openai, which speaks the Chat Completions format, with the base URL https://api.unbiased.ai/v1. That is the setup we ran with Qwen Code 0.25.3 on October 20, 2026: it found the off-by-one in a failing test, fixed it and reran the test to green. Format compatibility is not a promise that every feature works, so tell us if one doesn’t.

Why isn’t pareto in the model list?

Unbiased does not serve a model list, so Qwen Code cannot discover pareto on its own. Type it in OPENAI_MODEL, or declare it as an id under modelProviders; model.name must match that id exactly.

Why use UNBIASED_API_KEY in the settings file?

Exporting OPENAI_API_KEY works, but it is also the name other OpenAI tools read. The envKey field lets Qwen Code take your Unbiased key from its own variable and leave OPENAI_API_KEY alone.

Can I switch back?

Yes. Unset the three variables, or remove the modelProviders entry and selectedType, and run /auth to choose another provider.

Is Pareto made by the Qwen Code team?

No. Pareto is Unbiased’s model. Qwen Code is made by the Qwen team at Alibaba Cloud, and Unbiased is not affiliated with or endorsed by them. You keep Qwen Code and point it at the Unbiased API.

What happens when I hit the weekly allowance?

Requests over your plan’s limits return an HTTP 429 error. Personal includes 50M tokens a week and 20 requests per minute. Personal Max raises that to 600M a week and 60 requests per minute for $100/month. Team uses prepaid pay-as-you-go credits with no weekly token cap.

Do I choose between models?

No. You send every request to one model string, pareto. Pareto selects frontier and open models for each request and blends their outputs into one answer.

Is my code kept?

Personal and Personal Max are not our zero data retention tier: prompts and responses may be kept up to 30 days, as section 5(c) of the Terms explains. For work that must not be kept, ask about the Zero Data Retention API on the Private AI page.

Pareto works in other tools too

The same key and the same model string go into each of these. See them all.

Product names belong to their owners. Except where a guide says otherwise, Unbiased is not affiliated with or endorsed by the makers of the tools on this page.

Hand Qwen Code the test that keeps failing

Three variables, one key and the terminal you already use. Pareto starts at $10 a month for 50M tokens a week.

Choose Personal · $10/month See the setup →