Kai Mak

CRO @ Together AI

[Together AI @togethercompute]

DeepSeek V4 Pro 0813 is live on Together AI

DeepSeek’s flagship V4 Pro release brings a 1.6T MoE architecture, 1M context, and three reasoning modes for coding, agents, and complex reasoning.

[Vipul Ved Prakash @vipulved]

It has been clear to many of us, and now it’s becoming clear more tangibly, that AI models will commoditize to various degrees. This is of course a difficult business reality if your core business depends on exclusivity on intelligence.

But commodity markets are not communism. They are the largest markets on earth. Oil, grain, steel, electricity, memory: trillions clear through them every year, priced by competition among thousands of suppliers. In economic terms, communism is one provider and no price. A commodity market is the precise inverse. The world that actually resembles central planning is the one Dean argues for: a set of protected incumbents, access gated by the state, agencies instructed to manufacture FUD until every regulated buyer, and transitively every tool maker upstream, backs away from cheaper competitors.

Open weights don't deter capex. They move it. When the model layer commoditizes, spend shifts to inference, data, tooling, and applications, and builds far broader industrial infrastructure rather than concentrating capital in a handful of companies. Most of our digital infrastructure today, hyperscalers included, runs on open source. The businesses built atop it keep excellent margins and compound at extraordinary rates. Open-weights intelligence will likely rank among the most important economic accelerations in history. It won't be kind to every early incumbent, Linux wasn't kind to Sun Microsystems, but it will be very good for almost everyone else. I suspect OpenAI and Anthropic, given their positions, excellent products, resources and talent density, will be just fine. They will simply hold a little less pricing power.

The security theater around Mythos continues to do damage. Of course, there is no evidence for the hysterical claims. The evidence is in fact so thin that proponents of AI's existential risks now openly recommend FUD as the strategy. That should be telling.

[Aravind Srinivas @AravSrinivas]

Strong results

[Tony Zhao @tonyzzhao]

Introducing ACT-2 Preview

The first robotics model to unify broad generalization with high reliability. A single fine-tuning example can teach Memo a new behavior that generalizes.

Zero shot, real unseen homes, 99% success rate.

[Nathan Tindall @ntindall_]

Consistently impressed by the horsepower of @togethercompute and the people behind it! Thank you for your partnership!

[Together AI @togethercompute]

GLM 5.2 on Together AI is now #1 on Artificial Analysis for both output speed and latency.

[Vipul Ved Prakash @vipulved]

We @togethercompute believe intelligence should be abundant, not expensive.

Today we announced our Series C funding of $800m @ $8.3B valuation, to continue to build the world's most efficient platform for generative AI.

Thanks @nikogallogly for telling our story in @nytimes!

https://t.co/LGQ5t6rzya

[Together AI @togethercompute]

Multi-GPU kernels are the real test for coding models. Today at @aiDotEngineer, @simran_s_arora shared ParallelKernelBench, an open-source benchmark for evaluating whether LLMs can write fast CUDA kernels for real communication-heavy workloads.

Proud to see this work from the Together AI Frontier Performance team.

[Vipul Ved Prakash @vipulved]

https://t.co/TIeuZQUj5D

[Together AI @togethercompute]

🤖OpenAI's open models are here.

gpt-oss models just landed on Together AI. Achieves near-parity with o4- mini, trained using o3 techniques.

Build anything, deploy anywhere🔥

[Vipul Ved Prakash @vipulved]

The first @togethercompute GB200 cluster CDUs imbibing coolant in prep to go live next week! Each rack here is 1.4 exaflops of inference performance!

[Michael Lingelbach @mjlbach]

There's a reason so many startups, including ours, are choosing to train and deploy models with @togethercompute

Over the last 4 weeks, we spent countless hours getting hands on support from some of the best researchers and engineers in the world. We had staging up within a few days of handing off the model inference code, and soon were getting pull requests into our inference stack fixing bugs and massively reducing our time to market (sorry @realDanFu, I owe you a few weekends).

Last night at ~7 PM, I DM'd the team on our joint slack channel and told them to rollout a new pod with a fix for our latest model. Within 30 minutes we had hotfixed an issue that was causing approximately 20% of our generations to fail.

I wouldn't expect that from another company, but @vipulved has built an insanely world class team of engineers. Can't imagine better partners to build with.

[Vipul Ved Prakash @vipulved]

Rolling out a new inference stack for DeepSeek R1 @togethercompute that gets up to 110 t/s on the 671B parameter model!

[Together AI @togethercompute]

Honored that @karpathy used Together AI to use DeepSeek R1 in his latest video!

[Together AI @togethercompute]

Congratulations to @pika_labs on launching Pika 2.0! 🔥📹

We've been proud to support their team – from training on Together GPU clusters, to hosting, optimizing and scaling their incredible video models with Custom Model Inference.

Can't wait to see what creators build with this incredible model! 🚀

[Together AI @togethercompute]

FLUX has arrived on Together AI, and it's faster and more powerful than ever.

We’re one of the exclusive launch partners for FLUX1.1 [pro], @bfl_ml latest premium high-performance model.

Plus we’re giving developers 3 months free access to FLUX.1 [schnell] via our FLUX-schnell-Free endpoint. Start building with state-of-the-art image generation today.

Read more: https://t.co/PLbyBP3doR

[brryant @bryantchou]

We don’t talk about Webflow hosting often

- no ♾️ build steps, just one click.
- CDN on over 140 pops
- responsive image 🌅 optimization and more on the horizon
- we power ~1% of the internet (wild since our customers are larger businesses/premium)

🤩 to our infra teams

[Allan Leinwand @leinwand]

Don't ship on a Friday afternoon they said.
It's been in prod for a few days already they said 😅

Keyboard shortcuts FTW 🙌

[Webflow @webflow]

We're excited to announce and welcome our new Chief Technology Officer, @leinwand! 🥳

We spoke to Allan about everything from the biggest technical challenges in the industry, to being customer-obsessed, to what excites him about the tech space, and much more:

https://t.co/fCyrWOqQOb

[brryant @bryantchou]

We’re launching one of the first projects today out of Webflow Labs: a Figma to Webflow plugin!

It’s early, but in the spirit of moving quickly we’re excited to get your feedback on it.

Check it out on PH!

https://t.co/76iSZ1MfCi