Providers

Connect 20+ language model providers with your own keys, or stay fully local. Keys live in this browser session only.

OpenAI

Realtime voice and tool calling

cloud
Models
gpt-4.1gpt-4o-mini

Anthropic Claude

Long context reasoning

cloud
Models
claude-sonnet-4.5claude-haiku-4.5

DeepSeek

Low cost reasoning

cloud
Models
deepseek-chatdeepseek-reasoner

Google Gemini

Multimodal input

cloud
Models
gemini-2.5-flashgemini-2.5-pro

Mistral

European hosting

cloud
Models
mistral-largemistral-small

Groq

Fastest token stream

cloud
Models
llama-3.3-70bqwen-2.5-32b

xAI Grok

Live web awareness

cloud
Models
grok-4grok-4-mini

OpenRouter

One key, many models

cloud
Models
auto

Together

Open weight hosting

cloud
Models
llama-3.1-405b

Fireworks

Serverless open models

cloud
Models
qwen3-235b

Perplexity

Search grounded answers

cloud
Models
sonar-pro

Cohere

Retrieval focused

cloud
Models
command-a

WebGPU in browser

Runs fully on your device

local
Models
qwen3-1.7bllama-3.2-3b

Ollama

Local daemon on localhost

local
Models
llama3.2qwen2.5

LM Studio

OpenAI compatible server

local
Models
any loaded model

llama.cpp server

Bring your own weights

local
Models
gguf models

vLLM

High throughput self hosting

local
Models
self hosted