Launched this week
Osaurus
Open source agents that run 100% locally on your Mac
1K followers
Open source agents that run 100% locally on your Mac
1K followers
The native macOS harness for AI agents. Any model, persistent memory, autonomous execution, cryptographic identity. Built in Swift. Fully offline. Open source.
Mac-native and local is the combo I keep wanting, the Firefox-of-local-AI pitch landed. Practical question: what's the real memory footprint running one of the mid-size models on, say, a 16GB machine, does it stay usable or does it eat everything and make me quit my other apps? That's usually where "runs locally" turns into "runs locally if you bought the maxed-out box." Nice to see something leaning open source instead of another closed wrapper.
Osaurus
@chielephant I would say 16GB is going to be a little tight, but if you use something like Gemma 4 E4B, you should be able to use it alongside other apps.
The approval gate + local memory combo is the interesting bit for me. I’ve been nervous about agents touching real files; seeing every action before it runs feels like the right tradeoff. Curious how you handle rollback when an agent writes bad files?
Osaurus
@xiaosong001 You can ask the agent to undo the file operation, and it keeps an internal log of changes that it can revert to
Hi, this is cool. What are some exciting and unique ways that people are using Osaurus and is there a roadmap?
Osaurus
Hey @andrea_j , great question!
Honestly the sandbox stuff surprises me the most. Agents spin up an isolated Linux VM and make real things: slide decks, 3D games, deployed websites. Recently I watched an agent write its own skill to solve a task, which was a little surreal.
Roadmap: personal AI you actually own. Agents, memory, skills that live on your Mac and compound over time. Lots left to build.
Curious what you'd want your Mac to do on its own!
@tpae Are the latest local models to download updated regularly? How do I know which ones work well on my machine? What's the best way to find ones that meet my needs? Thanks
Osaurus
@andrea_j great question! We add new models very frequently, we quantize them ourselves for the best experience and compatibility with the harness. We let you know which models are too big, or which models are just the right fit.
@tpae Thanks - I'm starting to play with Osaurus atm. Could I give my dino agent a set of pdf books to read and he could talk/speak to me about them as an expert on those topics? Could he then research these topics daily and build a memory? Oh, and what makes this different than LLM Studio/ Ollama?
@Osaurus Congrats on the Launch. The Osaurus dino is super cute.
I have been waiting to try locally run models and see the performance. What the hardware requirement? is a M4 Mac mini 16gb enough? Eager to try it.
Osaurus
Hey @roopesh_donde !
16gb would be usable, although 24gb would be recommended. I would suggest using Gemma 4 e4b, it's not crazy smart like the others but is able to call tools in a meaningful way. Let me know if you run into any questions!
Does it take a lot of space in my computer?
Osaurus
@doganakbulut Depends on which model you download, but it ranges between 6GB ~ 20GB for smaller models, and could be more than 200GB+ for larger ones
DiffSense
Third and final question 🙏 Does Osaurus have a cli input? Like is it possible to invoke an action via cli to it? If you know what I mean?
Osaurus
@conduit_design Osaurus does have CLI input! You can read more about it here: https://docs.osaurus.ai/cli
You can invoke actions using our MCP server.
How is this different with Codex or Claude Code?
Osaurus
@jun_rex_bungabong Codex and Claude Code are designed for development, we're designed for every day use. We also focus on getting local models to work well, and provide local inferencing.