People of Pi: the next release will incorporate dynamic tool loading without cache wiping on supported models and providers. We did some investigations and found a way to get somewhat consistent API behavior between OpenAI and Anthropic. (thanks @zeeg for pushing)
very cool morning read
> Pi has slightly higher pass rate with Opus 4.8 than Claude Code, and sends less context (likely does better compaction)
> Pi with GLM 5.2 is better than GPT 5.5 and less costly than Opus + Pi
> same pattern with Codex too, just Opus 4.8 is a better model
gg @badlogicgames
At 11k employees, our AI costs are going up. Which model & harness should we use to lower cost but also retain great quality?
We didn't want to blindly trust public benchmarks. So we ran a comprehensive evaluation on our tasks, code base, infra. It's been produced by more than
3) Harnesses make a huge difference in cost-performance. The very simple Pi harness (@badlogicgames) got the same success rate as harnesses from the LLM vendors with Opus and GPT 5.5, but at 2x less cost! Seems to be mainly due to smaller inputs to the LLM.
People of pi we got an update for you. It can now show you in /session how much money you lost on cache misses, you can get cache warnings in transcript, GPT 5.6 is in there, new extension and lifecycle hooks, big update to "pi config" and loads of fixes.
People of pi, enjoy a little weekend update! /new with lots of extensions should now be slightly faster. Fixed GLM-5.2 configs on some providers, added a new way to configure reasoning on VLLM and others. edit tool's fuzzy algorithm is no longer touching non-modified lines.
Just Shipped: Flue 1.0 Beta
Flue is the TypeScript framework for building the next generation of agents, designed around an open agent harness with zero LLM lock-in. It’s like Astro, for agents.
Flue 1.0 has been redesigned around three core primitives:
🔁 Workflows —
People of Pi. To finish off the week, here is small update to fix a bunch of issues. Run pi update to get to 0.79.2. Among other things some fable fixes and better auto complete for /model.
People of pi we pushed out some changes. Fable is in there, slightly less annoying trust prompt and settings for it. And lots of other stuff. pi.dev/news/releases/…
People of pi: 0.79.0 is out and includes some security hardening. Most notably when you have an AGENTS.md file or .pi folder, we now ask you if you trust the folder before loading it. Global extensions can override this. pi.dev/news/releases/…
NVIDIA just released Nemotron 3 Ultra, a leading Hybrid Mamba-Transformer MoE open model for long-running agents.
I've been test driving it these past few days, letting it work on Pi issues and it performed amazingly well!
Update your pi.dev and use it via @OpenRouter today!
122 Followers 83 FollowingWeb3 builder & degen on @solana. On-chain vibes, off-chain impact. Forming @bullish4gency, strategist @kimcookalts and just vibin’ depends on the block.
1K Followers 689 Following1:1 Agentic Business Operations for Founders.
Polyagentic (Cloud and Local LLMs).
All-in on the goated combo: 🦞 + Pi + Opus/Codex.
136 Followers 312 FollowingOil Painter & Digital Concept Artist / Illustrator. Painting keyframes that sell the vision for film & commercials. Clients: Alpine, Audible, Rabanne, SAP, etc.
88 Followers 259 FollowingInvestor focusing on lvl 1 protocols and fundamentals
Best Tech + Store of value + Privacy + Decentralization
Not investment advice
#Decred🎟 #Bitcoin
14 Followers 164 Followingbuild products from concept to reality as they should be.
building https://t.co/VFdsVppDyK, proxypal, https://t.co/P4yLnpgZMO
259 Followers 442 FollowingHolding a PhD in hydrology, interested in riverine sediment transport issues and all things #geospatial. Love mountains, mystery stories and #rstats 🤎
560K Followers 2K FollowingPolyagentmorous ClawFather. Came back from retirement to mess with AI and help a lobster take over the world.
@OpenClaw🦞 + @OpenAI