Osaurus Studio

Plan your on-device AI harness: providers, local AI models, agents and exact API requests for your Mac.

Osaurus Studio — plan your on-device AI harness: providers, local AI models, agents and exact API requests for your Mac.

How to use Osaurus Studio

Start on Setup to check the requirements and install with brew install --cask osaurus
Pick a provider preset and generate the exact connection config and models test command
Browse the local model catalog, from bundled Raptor 0.6 to Gemma 4, Qwen3.6, GPT-OSS and Llama
Build a custom agent with persona, model, temperature and capabilities to get osaurus_config JSON
Use the API lab to build an exact chat request for the 127.0.0.1:1337 server and copy the curl
Check the feature matrix for what is Stable, Beta, Foundation or Experimental

Real Osaurus data from osaurus-ai/osaurus: provider presets, endpoints, CLI and feature matrix copied from the repo docs
Provider configurator with the exact presets: Anthropic, OpenAI, xAI, OpenRouter and Custom, plus Ollama and LM Studio notes
Local model catalog: Raptor 0.6, Gemma 4, Qwen3.6, GPT-OSS, Llama, plus image and voice models
Agent builder producing real osaurus_config JSON with delegation, budgets and capabilities
API lab generating exact curl and payloads for the 127.0.0.1:1337 OpenAI, Anthropic and Ollama endpoints
Honest scope: the native macOS app cannot run in a browser, so this studio plans and configures it

Yes, unlimited.