Warp, with Pareto behind the agent

Warp’s agent already lives where your builds fail: in the terminal, reading the same output you are. Keep the blocks, the input editor and the approvals; change only who does the thinking. Add Unbiased as a custom endpoint, pick pareto in the model picker, and Pareto selects frontier and open models for each request and blends their outputs into one answer.

Choose Personal · $10/month See the setup →

Settings from Warp’s docs50M tokens a week

Warp

npm test

FAIL src/slug.test.ts

Expected: "hello-world" Received: "hello-world-"

npm test is failing. Find the cause, fix it and rerun the tests.

Read src/slug.ts

- return s.toLowerCase().replace(/[^a-z0-9]+/g, "-");+ return s.toLowerCase().replace(/[^a-z0-9]+/g, "-").replace(/^-|-$/g, "");

Apply this edit? Accepted

Run npm test? Accepted

PASS src/slug.test.ts

Fixed. Titles ending in punctuation left a dash at the end of the slug. slugify() now trims leading and trailing dashes, and npm test passes.

model paretocustom endpoint · api.unbiased.ai

Ask the agent, or run a command

An illustration of Warp on Pareto. Not a screenshot.

Add Unbiased as a custom endpoint

Warp lets the Warp Agent use any endpoint that implements the OpenAI Chat Completions API at a public URL. Unbiased does, at https://api.unbiased.ai/v1. Warp sends your agent’s requests to it from Warp’s servers, so the address must be reachable on the internet, which this one is.

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.
Open the endpoint settings
In Warp, open Settings and search for inference endpoint. The custom inference settings sit under Agents > Warp Agent.
Enter the endpoint
For the endpoint URL, enter https://api.unbiased.ai/v1: the base URL, not the full /chat/completions path. Add your Unbiased key as the API key. Set the schema to OpenAI Chat Completions.
Type the model ID by hand
Enter pareto as the model ID and save. Unbiased does not serve a model list (GET /v1/models), so Warp will not discover it on its own.
Pick pareto, not Auto
Choose pareto in the Warp Agent’s model picker. Auto always uses Warp’s own inference, even with an endpoint configured.

Settings follow Warp’s own documentation: Warp docs: Custom inference endpoint.

What to enter in Warp

# Settings > search "inference endpoint"
Endpoint URL: https://api.unbiased.ai/v1
API key:      your-unbiased-api-key
Model ID:     pareto
# Then select pareto in the Warp Agent model picker (not Auto)

Warp’s changelog (2026.09.02) says endpoint definitions are shared by the Warp app and the Warp Agent CLI, so the same entry should appear under /model in the CLI. The CLI still needs you signed in to a Warp account.

Warp stays. The model changes.

The agent’s model changes. The terminal around it does not.

Blocks, the input editor and Agent modeUnchanged
Approvals for commands and editsUnchanged
Your Warp accountStill required
Auto and custom routersCan’t select pareto
Model in the pickerpareto
Where agent requests goapi.unbiased.ai/v1, through Warp

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 Warp 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 the agent’s model by the month

One API key works across Warp 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.

Warp on Pareto, questions answered

Does Warp work with a model outside its own list?

Warp documents custom inference endpoints for the Warp Agent, and any endpoint that implements the OpenAI Chat Completions API qualifies. Unbiased serves that format. We have not run Warp against Pareto ourselves, and format compatibility is not a promise that every feature works, so tell us if one doesn’t.

Why is the agent still using Warp’s models?

Check the model picker. Auto always uses Warp-provided inference and draws on your Warp usage, even when an endpoint is set up. Custom routers can’t point at endpoint models either. Pick pareto by name.

Do I still need a Warp plan?

You need a Warp account. Personal endpoints are available on Warp’s Free plan and eligible paid plans, for individuals and organizations with 10 or fewer employees; larger organizations need Business or Enterprise, where local runs on an endpoint still carry Warp platform charges. Unbiased bills the inference. Warp can’t enforce zero data retention for endpoint requests, so retention follows Unbiased’s terms.

Does it work for cloud agents and the Warp Agent CLI?

Personal endpoints cover local Warp Agent runs only, not cloud agents. Warp’s changelog says endpoint definitions are shared by the app and the Warp Agent CLI, where you choose a model with /model; the CLI still asks you to sign in to Warp. Warp’s other AI features keep their own model settings.

Can I switch back?

Yes. Pick a Warp-provided model, or Auto, in the model picker. Remove pareto from the endpoint settings if you want it out of the picker.

Is Pareto made by the Warp team?

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

Let Warp’s agent read the next red build

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