Introducing Claude Haiku 5.5: the cheapest, fastest, and most capable small model we’ve ever released.
On average, it costs around 75% less to run than Claude Haiku 4.5.
Decisions API, for lightning fast constrained decision making powered by Luna. Supports visual inputs, and tuned to be able to make decisions in less than a few hundreds of milliseconds end to end.
@sahildarz Current models are already good enough for most day-to-day work. I agree that the harness plays a crucial role in getting consistent, high-quality output.
I think Claude ✳️ is finally back! I was struggling to understand the text generated by Opus 5 before, but now it seems fixed and is delivering much better-quality code.
https://t.co/am0qaEWLnf
Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family.
It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.
Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family.
It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.
"AI moderation is too slow for real-time chat."
Not anymore.
@typesafeai Jev judges every message in ~200ms, before anyone else sees it,for a fraction of what an LLM call costs.
https://t.co/uTxhhKaVJE
I built a real-time chat where Jev judges every message in before the room sees it.
❌ No keyword lists
✅ Rules are plain English
✅ Unsure? Held for a human
Built with Jev (@typesafeai) + @convex + @vercel
Try to break it 👇
https://t.co/P5HYDS6aJX
https://t.co/gGWn6XmPaA
Qwen3.8-27B at 144 tok/s on an M5 Max MacBook Pro ⚡
Meet Inco Splash: our open-source inference engine, built around the model and around Apple silicon.
Up to 3× the decode speed of Ollama, 2× oMLX, and almost 4× when an agent fans out into sub-agents.
We're adding support for AGENTS.md to Claude Code.
Starting today in version 2.1.277, if there is no CLAUDE.md in a folder, Claude will check for and use AGENTS.md.
You can toggle this behavior in /config.
@hazemomier@typesafeai This is just an experiment to showcase what Jev is capable of. A production use case would need more research, and real traffic to test against, before I could answer that precisely.