SN50 is REALLY FAST. And you don't have to just take our word for it. ⚡️ @SemiAnalysis independently benchmarked SN50 running @MiniMax_AI M2.7 at ~800 t/s. That's consistent with the 800+ t/s we demonstrated at @RaiseSummit, validated by @ArtificialAnlys. Fast inference just got another third-party stamp of approval.

What if you use @Claude to orchestrate while @MiniMax_AI on SambaNova handles the focused coding work? The result: faster development cycles and significantly reduced costs. See it at AI Week Italy. May 21 | 6:30 PM Register below ⬇️

💡 @MiniMax_AI M2.7 is an open-weight LLM built for serious dev work. It’s the first in MiniMax’s M-series to “self-evolve” via its own training + eval loop (agent harness optimization). Designed for complex coding, multi-agent systems, and pro-grade workflows. Learn more:

UK sovereign AI is becoming reality. Together with Argyll , we’ve launched a sovereign AI inference cloud built on SambaNova’s full-stack AI platform — designed for performance, efficiency, and control without the trade-offs of traditional infrastructure. As AI adoption scales, sovereignty can’t just be a label. It has to be demonstrated through infrastructure ownership, operational control, energy efficiency, and where intelligence is deployed. This deployment delivers: ⚡ High-performance AI inference 🌍 Renewable-powered infrastructure 🔋 ~10kW rack density with air-cooled systems 🏗️ Disaggregated architecture across UK data centers 🧠 SambaNova RDUs for efficient large-scale AI This is what the next generation of AI infrastructure looks like. Read more via DatacenterDynamics ⤵️

The next AI war won’t be about training models. It’ll be about: ➡️ Inference costs ➡️ Compute shortages ➡️ Scaling AI profitably Following Cerebras’ IPO, @RodrigoLiang joined @Bloomberg with @mattmiller1973 and @daniburgz to explain why inference could become the biggest business in tech. Enterprise AI demand is exploding. The infra race is just getting started. Watch here ⤵️

🚀 @MiniMax_AI M2.7 running fastest on SambaCloud Watch it build a browser-based OS in a single HTML file, complete with working apps like Snake, Paint, and a calculator. Try it yourself:

Nostalgia hit for the millennials: let's build a Snake Game 🐍 Watch what happens when you ask an AI model to build something. Three-tier memory moves models and activations across DDR, HBM, and on-chip SRAM to keep inference fast & efficient.

We're proud to partner with Argyll Data Development to bring sovereign AI infrastructure to the UK. Built on SambaNova RDU architecture, this is about delivering high-performance inference while keeping data, models, and operations local. Excited to see what comes next 🦾

Big moment for the AI infrastructure industry today. The demand for next-generation AI infra providers is becoming impossible to ignore as the market expands beyond a single dominant player. We're in a new phase — one driven by more innovation, more competition, and more ways to scale AI efficiently. 🚀

Sydney showed up 🇦🇺 Huge thanks to the founders, devs, builders, & partners who joined us for SambaHouse with Equinix & SCX. Great convos, live demos, & a room full of people building real AI products with open-source models & sovereign infrastructure.

AI didn’t stop at training models. Now the challenge is turning those models into real applications people actually use. @SumtiJairath talks about the shift from training to inference & why efficiency across compute & communication matters. 🎧 Listen to the @dcdnews here:

Coding agents spend most of their time executing, editing files, running tests, retrying failures, and iterating. That’s where speed matters 🚀 Our new Responses API support makes it easier to connect coding agents to fast, production-ready models on SambaCloud. Read more:

Some snapshots from SambaHouse in Munich. Founders, engineers, and operators all in one room talking about what sovereign AI in Europe actually looks like. Big thanks to everyone who joined us! 📸: @InfercomAI

ICYMI: Our team breaks down the real-world benefits, limits, and tradeoffs of in-context learning at scale 🦾 Many-shot prompting can improve performance, but only when the task, model, and example selection line up. Read more in our blog.

AI is hitting a new phase and infrastructure is at the center of it. Don't forget to join @RodrigoLiang at @WebSummit Vancouver as he breaks down what’s changing and what comes next. 🗓️ May 13 | 12:20–12:40 PM

Instead of treating prompts as static, Agentic Context Engineering (ACE) explores how context can evolve over time through generation, reflection, and curation. Read the paper ⬇️

Built differently for inference 🦾 RDUs are designed to keep execution streaming continuously across the system, from memory to compute to parallel execution at scale. 256 RDUs working together can generate thousands of tokens in parallel for modern AI workloads. Learn more:

What happens when you keep Claude for orchestration and hand focused coding tasks to @MiniMax_AI on SambaNova? Faster workflows and dramatically lower cost. Join us live at AI Week Italy to see it in action. May 21 | 6:30 PM Pre-register ⬇️

🚀 @MiniMax_ai M2.7 isn’t just another coding model. It was used to improve its own agent harness across 100+ rounds of iteration, helping drive major gains in coding, multi-agent collaboration, and real-world workflows. Try it now: