Skip to main content
FeaturedDaily

Topic

Open-weight models

AI models whose weights are released publicly for anyone to download, run and fine-tune.

Nvidia1mo ago

Nvidia to buy Hugging Face for $12.93bn

Confirmed 3 September. Nvidia's second-largest acquisition, and the open-model registry changes hands.

2 min read · 3 September 2026
DeepSeek2mo ago

DeepSeek's V4-Pro reaches general availability, then prices jump up to 1,100%

The brief: a quiet pricing-page edit on 12 August, then peak/off-peak billing from 16 August.

2 min read · 16 August 2026
Moonshot AI2mo ago

Moonshot publishes Kimi K3 weights: 2.8T parameters

The largest open-weight release ever, eleven days after the hosted launch.

2 min read · 27 July 2026
Open-weight models3mo ago

Meituan open-sources LongCat-2.0: 1.6T model, trained without Nvidia

The brief: a food-delivery giant gives away a trillion-parameter AI built entirely on Chinese chips — and the benchmarks are its own.

2 min read · 29 June 2026
OpenAI3mo ago

The end of 'tokenmaxxing': enterprises ration AI spend

The brief: buyers are capping AI budgets and routing spend to cheaper Chinese open-weight models, just as OpenAI and Anthropic head for IPOs.

2 min read · 26 June 2026
DeepSeek4mo ago

The 2026 open-weight surge, in brief

The year free models caught the frontier. The brief.

2 min read · 14 June 2026
Alibaba (Qwen)4mo ago

Qwen: the open-source model juggernaut, explained

Alibaba's open-model juggernaut — what it is, why it dominates, and the 2026 shift.

2 min read · 13 June 2026
Moonshot AI4mo ago

Moonshot ships Kimi K2.7-Code — open weights, coding-focused

Another open model from China, and this one's actually downloadable. The brief.

2 min read · 12 June 2026
Mistral AI4mo ago

Mistral in talks to raise ~€3bn at a ~€20bn valuation

Europe's AI champion, repriced. The brief.

2 min read · 12 June 2026
MiniMax4mo ago

MiniMax ships M3 — open-weight frontier claims, weights to follow

Cheap, capable, and not quite downloadable yet. The brief on what shipped and what to watch.

2 min read · 1 June 2026
DeepSeek5mo ago

DeepSeek ships V4: 1.6T open-weight model, $1.74 a million tokens

The brief: two MIT-licensed MoE models, 1M context, and the first DeepSeek architecture built for Huawei chips.

2 min read · 24 April 2026
Mistral AI6mo ago

Mistral open-sources Voxtral: a top-tier text-to-speech model you can run yourself

Open weights. One GPU. Aimed at ElevenLabs. The brief.

2 min read · 26 March 2026