Get Axle
Your models. Your call.

Intelligence is a choice.
Keep yours open.

Some tasks need a different way of thinking. Bring the models you trust into one familiar place to build — including a provider we do not know yet, as long as it speaks the OpenAI API.
Connect a provider
your-next-project
01 / Choose your intelligence

Different models.
Same place to think.

Explore a few of the providers you can connect. The model changes. Your project stays yours.

Anthropic

Your provider account

Model family

Claude

Bring Claude into your workflow for reasoning, writing code, and working through changes.

Connect Anthropic

No request is sent in this preview.

Interactive preview · Select a provider to explore

A wider perspective

The right model is your decision.

A direct connection to your provider. Familiar tools around it. Model availability follows your account and each provider's current catalog.

Cloud

OpenAI

GPT & reasoning models

Work with OpenAI models using your API key.

Setup guide
Cloud

Anthropic

Claude

Bring Claude into your workflow for reasoning, writing code, and working through changes.

Setup guide
Cloud

Google

Gemini

Connect a Gemini API key and choose the model available in your Google account.

Setup guide
Cloud

OpenRouter

A wider model catalog

Access multiple model families through one provider account and one API key.

Setup guide
Cloud

Fireworks

Hosted open models

Use open model families hosted by Fireworks without running your own infrastructure.

Setup guide
Cloud

Cerebras

Fast inference

Connect to Cerebras inference with the model and account that fit your work.

Setup guide
Cloud

Groq

Hosted open models

Bring your Groq API key for an alternative route to supported open models.

Setup guide
Cloud

Mistral

Mistral & Codestral

Connect a Mistral API key to work with the Mistral and Codestral models available to you.

Setup guide
Cloud

DeepSeek

Reasoning and code

Use a DeepSeek API key for chat and coding models, including the reasoning variants.

Setup guide
Cloud

xAI

Grok

Add an xAI API key to bring the Grok models into the same model list.

Setup guide
Cloud

Ollama Cloud

Open models in the cloud

Use hosted Ollama models when a task needs more than your local hardware.

Setup guide
Local

Ollama

Your local models

Connect a local Ollama runtime. Your machine supplies the compute; no provider key is required.

Setup guide
Cloud

Your own endpoint

Any OpenAI-compatible API

Bring a base URL and an API key. Axle checks the provider, lists the models it serves, and lets you switch between them.

Setup guide
Same workspace. Different engine.

Nothing between you
and your next idea.

Choose a provider in Settings, add your key, and pick a model. Your projects and tools stay close at hand.

Read the model guide

Your account. Your usage.

Connect directly with your own provider key. Model costs are billed by your provider.

Bring the base URL, not just the name.

Paste any OpenAI-compatible endpoint with its API key. Axle identifies the provider, checks that it answers, then lists the models it serves.

Room to change your mind.

Pick the provider and model for your next task from the same workspace.

Local is an option, too.

Connect Ollama on your machine or use a compatible endpoint you control.

Provider keys you controlLocal inference supportedNo single model required

Less between you and your next idea

Good tools leave you a choice.

Connect the intelligence that works for you. Build from there.

Read the docs

Also available for Windows and Linux.