Skip to content
The Chat Agent
Six providers. Zero lock-in.

Every provider. One interface.

Connect OpenAI, Anthropic, Azure, Google Gemini, Ollama, or any OpenAI-compatible endpoint in minutes. Switch models mid-conversation, stream responses, and let models call tools — all from a single unified UI.

Connect a providerProvider setup docs

Supported providers

Six integrations, one connection panel

Every provider ships with streaming, tool-calling support where available, and encrypted key storage. Add as many as you need and switch between them freely.

OpenAI

Popular

The widely used API powering GPT-4o and GPT-4 Turbo. Strong reasoning, parallel function calling, and multimodal vision inputs — straight out of the box.

gpt-4ogpt-4o-minigpt-4-turbo
  • Streaming
  • Tool calling
  • Vision

Anthropic

Popular

Claude models excel at nuanced reasoning and long-horizon tasks. A 200 K-token context window and reliable instruction-following make them ideal for complex agentic workflows.

claude-opus-4-5claude-sonnet-4-5claude-haiku-4-5
  • Streaming
  • Tool calling
  • Vision
  • Long context

Azure OpenAI

Azure-hosted OpenAI deployments. Use your own deployment names, keep data inside a VNet, support region-specific deployments, and manage access via Azure's own access controls.

gpt-4o (deployment)gpt-4-turbo (deployment)
  • Streaming
  • Tool calling
  • Custom deployment URLs

Google Gemini

Gemini 1.5 Pro and Flash from Google DeepMind. Up to one million tokens of context, native multimodal understanding, and cost-effective inference with Flash.

gemini-1.5-progemini-1.5-flash
  • Streaming
  • Tool calling
  • Vision
  • Very long context

Ollama

Connect a reachable Ollama endpoint by providing its base URL and an optional API key. Use Llama 3, Mistral, Phi-3, Qwen, and more through a familiar OpenAI-compatible interface.

llama3mistralphi3qwen2
  • Streaming
  • Custom endpoint URL
  • OpenAI-compatible API

OpenAI-compatible

Any server that speaks the OpenAI API format — OpenRouter, Together AI, Groq, Fireworks AI, and more. Provide a reachable endpoint URL and an optional API key and you are ready.

OpenRouterTogether AIGroqFireworks AI
  • Streaming
  • Configurable base URL

Capability matrix

What each provider supports

Use this table to pick the right provider for your use case — or mix and match by connecting multiple providers at once.

ProviderStreamingTool callingVisionCustom endpointLong context
OpenAI
Anthropic
Azure OpenAI
Google Gemini
Ollama
OpenAI-compatible

Quick setup

Connect any provider in seconds

No configuration files or environment variables. Connect your providers directly in the app and your models appear immediately.

  1. 1

    Open Settings → Providers

    Navigate to your account settings and select the Providers tab to see all available integrations.

  2. 2

    Add a provider and paste your API key

    Click Add provider, choose your integration, and paste your API key. For Ollama or OpenAI-compatible endpoints, enter the reachable base URL and an optional key.

  3. 3

    Your models appear immediately

    The model picker updates in real time. Switch providers or models mid-conversation — your full chat context carries over seamlessly.

The providers settings panel — add a provider, paste your API key, and your models appear immediately.

Mix providers and switch mid-conversation

You can connect multiple providers simultaneously and switch between models at any point in a conversation. Open the model picker, choose a different provider or model, and your full chat context carries over seamlessly — no restart required.

Start building with The Chat Agent

Connect your favourite AI providers, add tools and agents, and ship in minutes. Free to start — no credit card required.

Get started freeRead provider docs