unbiased ai

Zed, with Pareto in the Agent Panel

You picked Zed for speed and you can keep it. Add Unbiased as an OpenAI-compatible provider, declare one model named pareto, and the Agent Panel and Inline Assistant can send their requests to Pareto. On each request Pareto selects frontier and open models and blends their outputs into one answer, while the editor itself stays as quick as it was.

Choose Personal · $10/month See the setup →

Settings from Zed’s docs50M tokens a week

An illustration of Zed on Pareto. Not a screenshot.

Add Pareto as a provider

Zed lets you add any OpenAI-compatible endpoint as a language model provider. You can do it from the Agent settings or in settings.json; both end in the same place.

  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. Open the LLM Providers settings

    Run agent: open settings from the command palette, which opens Settings → AI → LLM Providers. Click Add Provider and enter unbiased as the name, https://api.unbiased.ai/v1 as the API URL, pareto as the model ID and 1000000 as the context window.

  3. Or write it in settings.json

    Add the block on the right under language_models.openai_compatible. Zed’s OpenAI-compatible provider uses the models you declare, so the available_models entry is how pareto shows up in the picker.

  4. Give Zed your key

    Paste the key into the provider’s API key field, which Zed keeps in your system keychain, or set UNBIASED_API_KEY before launching Zed. Keep the key out of settings.json.

  5. Pick Pareto in the Agent Panel

    Open the model selector in the Agent Panel and choose Pareto under the unbiased provider. Ask it to do something small first.

Settings follow Zed’s own documentation: Zed docs: Use API Access, OpenAI-Compatible Endpoints.

~/.config/zed/settings.json
{
  "language_models": {
    "openai_compatible": {
      "unbiased": {
        "api_url": "https://api.unbiased.ai/v1",
        "available_models": [
          {
            "name": "pareto",
            "display_name": "Pareto",
            "max_tokens": 1000000,
            "max_output_tokens": 131072
          }
        ]
      }
    }
  }
}
Or set the key in your shell
# Zed builds the variable name from the provider ID
export UNBIASED_API_KEY=your-unbiased-api-key

On Windows the settings file lives at %APPDATA%\Zed\settings.json. A non-empty UNBIASED_API_KEY wins over a key saved in the keychain; unset it and restart Zed to stop using it.

Zed stays. The model changes.

One provider entry and one model name are new. The editor you tuned stays put.

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 Zed 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 Zed by the month

One API key works across Zed 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.

Zed on Pareto, questions answered

Which provider type should I choose in Zed?

Use the OpenAI-compatible provider with the API URL https://api.unbiased.ai/v1. Zed sends Chat Completions requests by default, which Unbiased serves; leave chat_completions at its default. Zed also has an Anthropic-compatible provider type, and Unbiased speaks that format at https://api.unbiased.ai, but you only need one.

Why doesn’t Pareto appear in the model list on its own?

Unbiased does not publish a model list endpoint, and Zed’s OpenAI-compatible provider expects you to declare models yourself. Add pareto under available_models (or as the model ID in Add Provider) and it appears in the picker as Pareto.

Does this change Edit Prediction or external agents in Zed?

No. The provider covers Zed’s own AI features: the Agent Panel, the Inline Assistant, commit messages and thread summaries. Edit Prediction has its own provider setting, and external agents such as Claude Agent or Codex running inside Zed use their own login and model settings.

Do I need a paid Zed plan?

Not for this. Requests to a provider you add with your own key go to that provider and are paid for by your Unbiased plan. To switch back, choose another model in the Agent Panel or remove the unbiased entry.

Has Unbiased tested Zed with Pareto?

Not in a recorded live run yet. The settings here follow Zed’s documentation. Format compatibility is not a promise that every feature works, so tell us if one doesn’t.

Is Pareto made by the Zed team?

No. Pareto is Unbiased’s model. Zed is made by Zed Industries, and Unbiased is not affiliated with or endorsed by them. You keep Zed 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.

Open the Agent Panel and hand it the failing test

One provider entry, one key, the same fast editor. Pareto starts at $10 a month for 50M tokens a week.

Choose Personal · $10/month See the setup →