Open WebUI, answering with Pareto

Q: Which connection type do I pick in Open WebUI?

An OpenAI API connection, under Manage OpenAI API Connections, with the URL https://api.unbiased.ai/v1. Open WebUI adds /chat/completions itself. Leave API Type on Chat Completions rather than Responses, which Open WebUI marks as experimental Open Responses support. Format compatibility is not a promise that every Open WebUI feature works, so tell us if one doesn’t.

Q: Why doesn’t pareto show up after I save the connection?

Open WebUI looks for models at /models, and Unbiased does not document that endpoint. Type pareto into the connection’s Model IDs list, click + and save; it then appears in the model selector. A failed Verify Connection check does not stop chat from working. If you can see it but your users can’t, set its access to Public under Settings > Admin > Models.

You host the chat app, your team keeps its workspaces, notes and chat history, and the answers come from Pareto. Add Unbiased as an OpenAI-compatible connection in the admin settings, type pareto into the Model IDs list, and it appears in the model selector. Set it to Public under Settings > Admin > Models and your whole team can pick it. Each reply comes from frontier and open models that Pareto selects and blends into one answer.

Choose Personal · $10/month See the setup →

Settings from Open WebUI’s docs50M tokens a week

Open WebUI

OIparetoOpen WebUI · api.unbiased.ai

Model: pareto

test_chunk.py fails with AssertionError: [[1, 2], [3, 4]]. What is wrong with chunk.py?

chunk.py attached

The loop stops one group early: for i in range(0, len(items) - size + 1, size):

With 5 items and a size of 2, that range yields 0 and 2. The slice that starts at 4 never runs, so the trailing 5 is dropped.

Loop to the end of the list instead: for i in range(0, len(items), size):

Changed it. python3 test_chunk.py prints ok.

Good. A slice that runs past the end of a list just stops there, so the last group needs no extra check.

Should I guard against a size of 0?

range() already raises ValueError for a step of 0. For a clearer message, check size < 1 at the top and raise your own.

How can I help you today?

Admin Settings › Connections

Edit Connection

URL
https://api.unbiased.ai/v1

Key
your-unbiased-api-key

Model IDs
pareto  +

API Type
Chat Completions

Saved. pareto is now in the model selector.

An illustration of Open WebUI on Pareto. Not a screenshot.

Add Pareto as a connection

Open WebUI talks to any provider that speaks the OpenAI Chat Completions format, and Unbiased does. An admin adds the connection once in the admin settings; it shows up for admins first, and you choose who else gets it under Settings > Admin > Models.

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 connection settings
As an admin, go to Settings > Admin > Connections (the menu reads Admin Panel > Settings > Connections), find Manage OpenAI API Connections and click + Add Connection.
Fill in the URL and key
Set URL to https://api.unbiased.ai/v1, keeping the /v1, and paste your key into API Key. Leave API Type on Chat Completions.
Add pareto by hand
Unbiased does not publish a model list, so type pareto under Model IDs, click +, then Save. If Verify Connection shows an error, chat can still work; Open WebUI’s docs say the same.
Pick pareto and ask
Start a new chat and choose pareto in the model selector. Use Set as default if you want every new chat to start there. To share it, open Settings > Admin > Models, edit pareto and set Access to Public or to the groups you choose.

Settings follow Open WebUI’s own documentation: Open WebUI docs: OpenAI-compatible providers.

Connection settings

URL:        https://api.unbiased.ai/v1
API Key:    your-unbiased-api-key
Model IDs:  pareto
API Type:   Chat Completions

Or set it at first launch (Docker)

# Read on first launch only
docker run -d -p 3000:8080 \
  -e OPENAI_API_BASE_URL=https://api.unbiased.ai/v1 \
  -e OPENAI_API_KEY=your-unbiased-api-key \
  -e WEBUI_SECRET_KEY=your-secret-key \
  -v open-webui:/app/backend/data \
  --name open-webui --restart always ghcr.io/open-webui/open-webui:main
# Then add pareto under Model IDs

Docker serves Open WebUI on http://localhost:3000; open-webui serve uses port 8080. The environment variables are persistent settings, so once the instance has started, change the connection in the admin panel instead.

Your instance stays. The answers change.

Pareto arrives as one more connection. Your users, chats and workspace settings are where you left them.

Users, chat history and workspacesUnchanged
Your other connections, Ollama includedUnchanged
Where Open WebUI runsYour server or laptop
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 Open WebUI 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

Open WebUI is a chat app, but the price of a finished task still says a lot about a model. The model card publishes both cost and success rate 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 answers by the month

Open WebUI is free to run. Pareto is the part you pay for: pick a monthly allowance, or pay per token on Team. The same key works in any other tool you point at Pareto.

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.

Open WebUI on Pareto, questions answered

Which connection type do I pick in Open WebUI?

An OpenAI API connection, under Manage OpenAI API Connections, with the URL https://api.unbiased.ai/v1. Open WebUI adds /chat/completions itself. Leave API Type on Chat Completions rather than Responses, which Open WebUI marks as experimental Open Responses support. Format compatibility is not a promise that every Open WebUI feature works, so tell us if one doesn’t.

Why doesn’t pareto show up after I save the connection?

Open WebUI looks for models at /models, and Unbiased does not document that endpoint. Type pareto into the connection’s Model IDs list, click + and save; it then appears in the model selector. A failed Verify Connection check does not stop chat from working. If you can see it but your users can’t, set its access to Public under Settings > Admin > Models.

I set OPENAI_API_BASE_URL, but my change didn’t stick. Why?

Open WebUI reads those variables on first launch and stores them. After that, the admin panel’s values win, so edit the connection there, or set ENABLE_PERSISTENT_CONFIG=False if you would rather the environment always rule.

Does each person on my instance need their own Unbiased key?

No. The key lives on the admin’s connection, and anyone the admin lets use pareto can choose it. Personal plans are for one person, though, so an instance shared by a team belongs on the Team plan.

Can I keep my local models as well?

Yes. Pareto is one connection among however many you have. Ollama models and other providers stay in the selector, and the connection’s toggle turns Pareto off without deleting it.

Is Pareto made by the Open WebUI team?

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

Add one model ID and ask it something hard

One URL, one key and the chat app you already host. Pareto starts at $10 a month for 50M tokens a week.