Claude Code, running on Pareto

Q: Can I switch back to Claude?

Yes. Unset the variables, or remove them from settings.json (values there override your shell), and Claude Code goes back to your Anthropic login. If you ran /logout, run /login again.

If usage limits or the bill are what’s wrong with Claude Code, keep the terminal agent, its slash commands and its habits, and change the model underneath. Point Claude Code at Unbiased with a few environment variables and every request goes to Pareto, which consults a blend of frontier and open models and sends back one answer.

Choose Personal · $10/month See the setup →

Tested live with Claude Code 2.1.30550M tokens a week

~/shop · claude

export ANTHROPIC_BASE_URL=https://api.unbiased.ai

export ANTHROPIC_MODEL=pareto

claude

test_chunk.py fails. Fix chunk.py and rerun the test.

Read(chunk.py)

Read 6 lines

Bash(python3 test_chunk.py)

AssertionError: [[1, 2], [3, 4]]

Update(chunk.py)

- for i in range(0, len(items) - size + 1, size):+ for i in range(0, len(items), size):

Bash(python3 test_chunk.py)

Fixed. The loop stopped one group early, so the trailing 5 was dropped. It now runs to the end of the list and the test passes.

model pareto7 turns · 11s

Try “refactor the cart totals”

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

Set it up with environment variables

Claude Code reads its model settings from the environment. Set these in your shell, or put them in the env block of ~/.claude/settings.json so every session picks them up.

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.
Point Claude Code at Unbiased
Set ANTHROPIC_BASE_URL to https://api.unbiased.ai and put your key in ANTHROPIC_AUTH_TOKEN. Leave /v1 off: Claude Code adds it.
Name the model pareto
Set ANTHROPIC_MODEL to pareto, and point the ANTHROPIC_DEFAULT_HAIKU_MODEL, _SONNET_ and _OPUS_MODEL aliases at it too, so subagents and /model choices stay on Pareto.
Start Claude Code
Run claude in any project. Claude Code may print a note that it does not recognize the model name; the session runs normally.

Tested on October 20, 2026Claude Code 2.1.305 on Pareto (pareto). It found the off-by-one in a failing Python test, edited the file and reran the test to green in 7 turns. Run record.

Shell

export ANTHROPIC_BASE_URL=https://api.unbiased.ai
export ANTHROPIC_AUTH_TOKEN=your-unbiased-api-key
export ANTHROPIC_MODEL=pareto
export ANTHROPIC_DEFAULT_HAIKU_MODEL=pareto
export ANTHROPIC_DEFAULT_SONNET_MODEL=pareto
export ANTHROPIC_DEFAULT_OPUS_MODEL=pareto
claude

Or ~/.claude/settings.json

{
  "env": {
    "ANTHROPIC_BASE_URL": "https://api.unbiased.ai",
    "ANTHROPIC_AUTH_TOKEN": "your-unbiased-api-key",
    "ANTHROPIC_MODEL": "pareto",
    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "pareto",
    "ANTHROPIC_DEFAULT_SONNET_MODEL": "pareto",
    "ANTHROPIC_DEFAULT_OPUS_MODEL": "pareto"
  }
}

If you also have ANTHROPIC_API_KEY set, unset it in this shell so Claude Code sends your Unbiased key.

Claude Code stays. The model changes.

The base URL and the model name change. Almost everything else stays as you left it.

Slash commands, hooks and CLAUDE.mdUnchanged
File edits, permissions and tool approvalsUnchanged
Remote Control and voice dictationOff (they need a claude.ai login)
Model stringpareto
Where requests goapi.unbiased.ai
What you payFrom $10 a month

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 Claude 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
DeepSWE v1.1Changes an agent makes in a real codebase: $0.24/task

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 Claude Code by the month

One API key works across Claude 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.

Claude Code on Pareto, questions answered

Does Claude Code work with a model that isn’t Claude?

Yes, in our test. Claude Code sends Anthropic Messages requests to whatever ANTHROPIC_BASE_URL names, and Unbiased accepts that format. Anthropic does not support Claude Code with non-Claude models, so for help with this setup, contact Unbiased rather than Anthropic. We ran Claude Code 2.1.305 against Pareto on October 20, 2026: it read a failing Python test, edited the file and reran the test to green. Format compatibility is not a promise that every feature works, so tell us if one doesn’t.

Why does Claude Code say it doesn’t recognize the model?

Claude Code knows Anthropic’s model names, and pareto is not one of them. It prints a note, assumes a 200K-token context window, and carries on; our test run showed the note and finished normally.

Do I need an Anthropic account or a Claude subscription?

No. With these settings Claude Code sends requests to Unbiased with your Unbiased key, and your Unbiased plan pays for them. Claude Code itself is free to install.

Can I switch back to Claude?

Yes. Unset the variables, or remove them from settings.json (values there override your shell), and Claude Code goes back to your Anthropic login. If you ran /logout, run /login again.

Is Pareto made by the Claude Code team?

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

Give Claude Code the bug you’ve been putting off

A few variables, one key and the same terminal. Pareto starts at $10 a month for 50M tokens a week.