connor retweeted
We launched Baseten on Microsoft Foundry at Ignite today.
Azure customers now have access to open model inference on Baseten within their existing governance and access controls on Foundry, with zero data retention, and billed under the Azure agreement they already have.
baseten.co/blog/baseten-is-n…
connor retweeted
In my ongoing quest to find the limits of Opus 5.5, I directed it to audit my old scientific field, theoretical astrophysics. We indexed about 150,000 papers, audited 50,000 for issues, and did progressively more thorough checks on thousands more. We found nearly 1000 major errors that got cited many times, most are still uncorrected. This survey enabled me to estimate the efficacy of peer review for correcting errors (not very good) and to identify dozens of open questions that anyone with spare Claude Max tokens could answer today (.md context files included).
Bottom line: AI can speed up quantitative science by >1000x with no further model improvement required. But I'm seeing no evidence of this being pushed within universities, only academics complaining about labs dropping math proofs.
Among my fellow reformed academic high agency founders, however, several are speed running entire disciplines.
This survey has implications for how science will be run in the very near future. Hint: Academic publishing is dead. It was already a very stupid way to record and promulgate knowledge, and had been so for decades, even before all that publicly funded knowledge was locked behind paywall cartels. Now AIs can generate thousands of novel, correct papers per day, it is no longer meaningful for humans to consider a research paper a "unit of science" or to even bother much with the 99.9% of science which, up to this point, was necessary drudgery. Time to discover something like General Relativity or Evolution every week. LFG!!
Enjoy: science.caseyhandmer.com/phy…
connor retweeted
> it took us 4 days and 500$ in tokens
> we beat SGLang by 15%
> it took 6 prompts
> 3 out of those 6 were:
(a) "carry on"
(b) "proceed"
(c) "push to github"
and... we did it from SCRATCH (git init)
context: it has recently become a very clear industry trend to build a custom inference engine for each open source model that comes out... a paradigm made popular by a WAVE of papers that came out earlier this year, like hazy research's megakernels and MetaInfer
the usual argument against traditional inference engines (like vllm / sglang) is that they, like all conventional swe products, use abstractions. the use of abstractions is synonymous with performance overhead (triton vs cuda, python vs c++, etc). combine this with increasing hardware complexity, increasing parallelization strategies, increasing kernel specialization techniques... and you can make a very clear (but very FLAWED) argument for building an engine customized purposely for one model.
why flawed? it is true that traditional inference engines waste cpu time running generic abstractions. sure. but the gpu kernels (the functions run on the gpu) are almost always longer, and so the wasted cpu time is not the bottleneck. if you know your shapes, and run cuda graphs ahead of time (and you almost always do), this is NEVER a problem.
so why would you build your own engine? why are people RAVING about this?
so we tested it out. we built one for DeepSeek V4.1 Flash. The goal was two-fold:
1) is it feasible to do so (spoiler: yes)
2) is it useful to do so (spoiler: also yes...but for different reasons)
we built it in rust (in retrospect, this did not really matter), and it took us 6 prompts and cost 500$ end to end.
unshackled from working within a pre-existing codebase or inheriting code debt, we:
- implemented custom kernels (GEMV, MoE, and AllReduces) specifically for DSV4.1 at our batch size
- we were able to put an entire spec dec step in one graph, and verify across 7 tokens at that
- we were able to mix deepseek's 4 residual streams in 1 fused kernel instead of 80
- fused the routing so that picking and running the expert happened in a single step (saving one write and one read i/o from each expert)
iterating on the model was cheap to do in a custom sandbox-like engine, and we were able to go much faster than a traditional dev cycle. we were then able to port every single thing into sglang, and matched performance
imo, it is this easy iteration that makes this paradigm useful. little else matters.
and, ofc, we've open-sourced the engine and upstream(ed/ing) the patches to @sgl_project
connor retweeted
Replying to @_catwu
@_catwu was the PM behind Claude Code. in 2 years, Claude won the hearts, minds, and tokens of the majority of developers on earth.
coding is the main character in this phase of AI. something else will come.. but we could not be where we are today without cat's contributions.
Friday was my last day at Anthropic.
Two years ago, when Claude Code was a small internal tool, I asked it to build an RL env. I steered it for tens of hours but never manually wrote a line of code. It worked, I was hooked, and I joined the team.
When we first launched, most people approved every permission prompt by hand and mostly used Claude to answer questions and write tests.
Today, the default is trusting Claude's judgment to auto-approve. Claude does large migrations and builds new products.
Helping build Claude Code into an industry-defining product has been the journey of a lifetime.
Thank you to the incredible team behind Claude Code, Cowork, and Tag. These products are the sum of thousands of great ideas from you. I'll miss working with you all.
Thank you to our users for embracing a new way of working. Your constant feedback shaped so much of what we shipped and your stories made the work so meaningful.
Excited to see what you all make next!
connor retweeted
The team is cooking today 🔥
The open frontier is built on @baseten and people are ready. Open-weight models have caught up to the frontier on coding, and our customers want to be able to pick the right model for each job, train it on their own data, and serve it with confidence at scale.
So this week we're shipping the building blocks for exactly that. Grid, our new gateway, router, and guardrails, launches tomorrow, Sandboxes launch on Thursday, and Loops, our serverless training, goes GA on Friday. @amiruci has the full rundown here:
connor retweeted
We're doing 3 launches this week, expanding Baseten's inference platform:
- Grid: gateway, intelligent router, and guardrails for coding agents (launching on Wednesday)
- Sandboxes: stateful, VM-isolated, and co-located with inference (launching on Thursday)
- Loops: serverless RL and SFT on frontier open-weight models (GA on Friday)
What excites me most is bringing training, inference, routing, and execution onto the same platform and showing what becomes possible when they're built to work together.
More to come all week.
connor retweeted
The best team all season left no doubt. Congrats to the @Brewers and everyone in Milwaukee on your first World Series title!
With open models, we can build safety tooling directly into the inference layer.
Our CEO and co-founder @tuhinone sat down with @kr00ney at @cnbc to discuss Project Beacon, and our work with @GoodfireAI to make safety the default for open intelligence.
@Tuhinone on CNBC
Today we’re announcing the Baseten Environment Network, a group of the best RL environment companies that will help our customers build custom environments for post-training.
The Network extends our applied research capacity, helping more customers build environments around their own tasks, data, and feedback.
Partners work directly with customers to create the environments, while we continue to handle model training and production deployment.
Learn more: baseten.co/blog/introducing-…
Today we release mammoth-1 and elephant-1!
We post-trained mammoth-1, the #1 ranked best decision model in the world, and elephant-1, which is the best model under 15B (and beats Jev).
We're open-sourcing all model weights and provide a technical report detailing all our training and data methods.
Today we’re releasing two decision models, elephant-1 (9B) and mammoth-1 (27B), trained by the @baselabs team, available for free for one week on our Model APIs.
mammoth-1 is #1 on the Jev Decision Index and elephant-1 is tied for #1 on the Artificial Analysis Decision Index.
It’s a new model, but more importantly, this is an open-source model you can access, deploy, and post-train on Baseten (or wherever you do your best work).
The free week ends 10/27/26 at 11:59 PM PT. Try it and let us know what you think.
Try mammoth-1 here: baseten.co/library/mammoth-1…
Try elephant-1 here:
baseten.co/library/elephant-…
connor retweeted
1/4 Today, we’re releasing Mercury 3. Mercury 3 is by far our most powerful model and the most intelligent diffusion LLM ever released. It outperforms GPT-6 Luna and Haiku 5.5 on quality while running at 1000+ tok/sec.
🤖 Made with AI
connor retweeted
There are literally zero moats left unless you go to the portfolio tab on our website those are the only companies with moats
Open models can come from anywhere. Securing them shouldn’t depend on where you get them.
Today we're launching Project Beacon to make safety a default for open intelligence.
We're partnering with @GoodfireAI to bring inference and in-line monitoring together for the first time, so safety signals are caught proactively.
Read more in our announcement: baseten.co/blog/safe-open-mo…
connor retweeted
We’ve lost an icon. Margaret Hamilton, a computer science pioneer best known for her work at MIT leading development of the flight software for the Apollo missions to the moon, has died at 90. Read about her legacy via MIT News: news.mit.edu/2026/margaret-h…





























