Building AI in the open makes the world more secure.
We’re joining the Open Secure AI Alliance to build open tools that safeguard software and agents.
We’ll keep releasing model weights and evaluations and sharing our research to strengthen a broader open ecosystem for defenders.
Excited to contribute alongside @nvidia and the other organizations building this ecosystem.
This has been a long time in the making and I'm so happy to see it out in the world.
From @almonk and @dizzyup's work before I joined @poolsideai through all the iterations as we've learned together what an AI coding tool should be.
It's been a real team effort https://t.co/e4gp7uifke https://t.co/YkR5AWKXRG @Johan_Lajili@pemsbr https://t.co/RmRIUpucL9 @mrncst@oyaaaaaaaasumi@glennui@prekesh and loads of others too🙏❤️
Today, we’re releasing Poolside Desktop Assistant.
One place to run coding agents across macOS, VS Code, and Visual Studio.
We built it for ourselves and have used it every day for the past year. Now we’re opening it up to everyone.
@antirez This is a result of trade-offs we made for efficiency and agentic capabilities, we'll take the feedback and see how it can be improved in future drops. Thanks for the push for Laguna on DwarfStar, exciting!
Still didn't attempt any quality testing, but let's say that GPT 5.6 Sol was able almost unassisted to write this Laguna S2.1 inference implementation following the other two models in DwarfStar, and it is very fast, 50 t/s generation, very fast prefill as well.
this is the drop the local ai crowd should be losing their minds over.
poolside just dropped laguna s 2.1: 118b total parameters, only 8b active per token, a full 1m context window, open weights under a real open license, on huggingface today.
look at the chart. it lands at 71 on terminal-bench at 118b, sitting above deepseek v4 pro max at a trillion params, above inkling at 1.5 trillion, above nemotron 3 ultra. it's beating models ten times its size and losing only to kimi k3, which is 24 times bigger. that's the efficiency frontier, up and to the left, exactly where you want a model to sit.
but here's the part that made me sit up: it runs on a single dgx spark.
and this is what nobody's saying loud enough. the dgx spark is the moe king. a dense 118b would crawl on it, the bandwidth chokes reading every weight each token. a moe with 8b active only ever reads 8b, so the spark's 128 gigs holds the whole model while generation stays fast. big brain, light footprint, the exact shape the spark was built to run.
open, frontier competitive, moe efficient, and it fits on a box on your desk. that's the whole thesis in one release: you don't need a datacenter, you need the right architecture on the right hardware. go grab the link below, weights are up.
Quite excited about this one: Laguna S 2.1: 118B-A8B MoE with 1M context. It runs on a single NVIDIA DGX Spark, and it stacks up very well on what we focus on: agentic coding.
A few observations we had building it. 🧵
Today we are releasing Laguna S 2.1.
At 118B total parameters, with 8B active per token, it does the work of models several times its size on agentic coding. It is remarkably persistent across long-horizon tasks. And it is small enough to run on a single NVIDIA DGX Spark.
It is far more capable than anything we have created before, and I think it redefines what a model in its weight class can do.
Laguna S 2.1 is an important model for Poolside. What it represents is even more important.
If, five years ago, I had read a book that said that by 2030 everything economically valuable, scientifically interesting, and personally meaningful would be built on intelligence contracted from three or four companies, I would have called it dystopian science fiction.
We are at a fork in the road of what kind of world we can have.
I believe intelligence should and will become a commodity. The question is whether that intelligence comes from three companies, or from many people who can build it, own it, and shape it.
The open ecosystem will not win by being the best in its own category. No one cares who is king of the open-source kingdom. People want the best intelligence for the task they are trying to do, with the right balance of quality, speed, cost, and control.
If we want a different future, open models have to be on par with, or better than, their closed equivalents.
Laguna S 2.1 is a meaningful step in that direction: capable enough to compete far above its weight class, efficient enough to run on hardware you can own, and open-weight so anyone can build on it.
Open-weighting our models is the contribution we can make today toward a world where intelligence can be built and owned by many. And we will keep doing it.
I am very proud of this team’s work. A big shout out to everyone at Poolside who made this possible, from infrastructure and data to architecture, pretraining, post-training, evaluations, and inference.
Laguna S 2.1 is available today under the OpenMDW-1.1 license, with weights on Hugging Face and access through OpenRouter and our API.
We are building toward a future where the most capable intelligence in the world can be owned and shaped by anyone. Laguna S 2.1 is one step. We are going to keep building until that future exists.
https://t.co/eTV7iNATrB
We're releasing ZML/LLMD, our homegrown LLM server built on top of our homegrown high performance heterogeneous inference stack.
It ships with 5 architectures out of the box: NVIDIA, AMD, Metal, Intel and TPU. All transparent.
It supports DFlash, continuous batching, prefix caching, the whole deal.
Oh, and it's fast.
We also worked on making XS 2.1 easier to run outside hosted endpoints.
XS 2.1 is supported in vLLM, SGLang, TensorRT-LLM, Hugging Face Transformers, and @ollama, with BF16, FP8, INT4, and NVFP4 checkpoints available.
We’re releasing DFlash speculator models alongside the checkpoints to speed up inference.
llama.cpp support and GGUFs are coming soon. We heard you.
https://t.co/RlfF2dWNd1