Almost every organisation on Earth is paying for the privilege to hand over its IP to someone else's AI.
We just closed a $50M Series A, led by @NEA, to build the alternative from right here in London: every model, agent and workload sealed in its own enclave, on infrastructure you control, with keys only you hold.
You can't be sovereign on someone else's stack. Build your own with Valarian.
AI models running inside critical systems can often reach far more data than they should.
@max_buchan, CEO of @Valarian, built their product to fix that, and how fast companies signed on shows just how real the problem already was.
Congrats to @valarian on its $50M Series A, led by @NEA. 🎉
The funding will accelerate the rollout of sovereign AI infrastructure, integrating frontier AI while protecting intellectual property, secrets and data from 3rd parties.
We implemented the harness with the Responses API and turned on:
→ Retained reasoning
→ Context compaction
On the public set, GPT-5.6 Sol’s score rose 188% while using 6x fewer output tokens.
AI security advances when the industry goes on holiday in August, together.
We're introducing the Open August Holiday Alliance and invite timelime leaders to develop new techniques and tools to safeguard their holiday and euro-maxx.
Learn more about joining the Alliance by dropping us a dm and we will add you as a signature:
Releasing the model weights and technical report of Kimi K3.
Kimi K3 is our most capable model: a 2.8T MoE model with native visual understanding and a 1M-token context window.
New model architecture: 2.5x the intelligence per unit of compute, not just more params.
Alongside Kimi K3, we're opening up more of the stack behind it — high-performance attention kernels, MoE communication library, and infrastructure for running agent environments at scale.
Model weights: https://t.co/7m7eEg6Y0B
Tech report: https://t.co/yeu6cjpMCT
Tech blog: https://t.co/YTfiMSNM1f
There's debate over whether it's "fully open" as it's not a standard OSI license like MIT or Apache.
But monetising at the top while remaining free and open for everyone else could prove to be a common path forward for open-weight models. https://t.co/g8xBsnjYau
Kimi K3 license. It's inspired by MIT but distinctly non-commercial, where any company making over $20M/yr must get a specific commercial deal (and display Kimi K3 if over 100M users or $20M/mo revenue)
The OWL is back 🦉
Matchday 02 of The Open-Weight League:
🇨🇳 Kimi K3 holds the top spot (provisional until its weights ship this week)
🇺🇸 Nemotron 3 Ultra stays the top American model at 3rd
🇪🇸 HyperNova 60B claims the top spot in the European division, passing Mistral
For my first post, I’m sharing a letter @NVIDIA signed on why open models matter.
AI will transform every industry, power every company, and be built by every country.
Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.
The world needs both frontier closed models and frontier open models.
https://t.co/AUKzoQ5Ikb