FastH3 v2 is now live on TensorScale! To celebrate the launch, we’re giving new users $10 in credits for a limited time, that’s up to 250 seconds of H3 video generation. Try it out at tensorscale.io Huge credits to @haoailab for making this high-quality, post-trained version of H3 available.

We’re hiring a Founding GTM Lead at TensorScale. Since launch, we’ve seen tremendous interest from developers and AI teams. Speed, cost, and ease of integration are consistently resonating with customers. Thank you to everyone who has tried TensorScale and shared feedback with us. What our early customers are saying: “Your endpoints are super fast, it’s really impressive.” “It is indeed very fast, an order of magnitude faster in fact!” “We have already integrated it into our app, and we’ve been surprised by how fast it works! The API docs are very clear, and we didn’t have any issues integrating it.” “I’m currently integrating TensorScale into my pipeline. The experience has been smooth, and I’m really happy with the results I’ve seen so far.” Now we’re looking for someone to help turn this momentum into growth — owning GTM strategy, positioning, customer acquisition, distribution, conversion, and expansion. Interested? DM me or apply here: tensorscale.io/careers-found…

@MiniMax_AI H3 ref-to-video is now live on @TensorScaleAI , with near-real-time inference This brings together Alibaba Cloud PAI’s distilled H3 model with TensorScale’s inference stack, achieving near-real-time speed: 17.7 seconds to generate 15-second 768p video for a real time factor (RTF) of 1.18. We’re continuing to optimize the stack and expect to push RTF below 1, enabling faster-than-real-time generation. Huge credit to MiniMax for the foundational model work, and to the Alibaba PAI team and broader open-source community for pushing the ecosystem forward. We’re excited to build on these efforts and make high-quality video generation faster, cheaper, and easier to serve in production.

RT @YaoZhangX: SenseNova U1.5 (@SenseTime_AI) is now live on @TensorScaleAI: frontier-quality image generation at 10x the speed and 60% of…

LTX 2.5 JUST WENT FASTER THAN REAL TIME LTX (@ltx_io) is already one of the world's most widely used open video models. We just made its latest release, LTX 2.5, fast enough for rapid iteration and affordable enough for production at scale.  It’s live on @TensorScaleAI now, generating a 5-sec 1080p video in just 3.12s at $0.04 per second. That's faster than real time, at a fraction of the cost! For creative teams, that means less waiting time, more ideas explored, and lower production costs. For robotics and physical AI teams, that means more data generated at scale, more scenarios tested, and shorter development and deployment cycles. Request early access today and we’ll get you onboarded! (link: lnkd.in/gV4tzeUK)

Today, we’re coming out of stealth with TensorScale (@TensorScaleAI ), and we're already running the fastest video inference in the world. MiniMax (@MiniMax_AI ) H3 — one of the leading open video models — now runs 10x faster and at half the cost on our inference stack. I’m building the company with long-time colleagues Teng Teng (@tt_tsio), Ye Wang (@yewang_ts), and Yun Tang (@_tangyun_). The AI stack was built for language models. We're rebuilding it for video and world models.