Blog
Product updates, engineering notes, and essays on local AI from the people building Osaurus.
Raptor 0.6: A Small Agent Model on a New Base
Raptor 0.5 shipped and then it looped. What we found, why we moved to Spark-X2.5-4B, and how the 3.7 GB Raptor 0.6 was built for the Osaurus harness.
Read postRaptor 0.5: A Small Agent Model for the Macs Everyone Actually Owns
An open 8B-A1B model tuned for automation and business work. 6.3 GB on disk, built to run on 8–16 GB Macs.
Read postAppleScript-8B & AppleScript-16B-A4B: Small Models That Actually Drive Your Mac
Two open, on-device models that turn plain-English requests into working macOS automation.
Read postThe Obvious Case for Open and Local AI
AI is a step change in what one person can do—but the power is being rented back to us. Here's the case for intelligence you can run, inspect, and keep.
Read postOn Decentralized Acceleration
Applying Vitalik's d/acc framework to AI — why every cloud prompt is a quiet transfer of power, and what the alternative looks like.
Read postOn Personal AI
A thesis on personal AI: why it should be private, local, composable, and owned by the people who use it—and how Osaurus is building that future.
Read postBuilding on the Edge: Why We're Betting on Local-First AI
What's happening with Dinoki, where Osaurus fits in, and why we're building AI infrastructure that puts you at the center.
Read postThe Missing macOS LLM Server
Why we built Osaurus: a native macOS LLM server for developers who want private local models, OpenAI-compatible APIs, and tools crafted for the Mac.
Read post
Get new posts by email.
Product updates and engineering notes, when we publish them. No account, and you can unsubscribe any time.