Palmyra X6 runs on GLM-5.2, priced $2 in and $8 out by stephane 15 August 2026 0 Read the fourth paragraph of the press release and the interesting sentence is right there, unhedged: Palmyra X6 was post-trained... Read moreDetails
Qwen3.8-27B ships Apache 2.0 with vision, the 2.4T doesn’t by stephane 15 August 2026 0 Open the two model cards side by side and the surprise is which one you would rather self host. Qwen3.8-27B... Read moreDetails
GLM-5.3: open weights in two weeks, breaking API change now by stephane 14 August 2026 0 Z.ai put a Hugging Face button on its own GLM-5.3 announcement page, and the button says Coming Soon. That is... Read moreDetails
GPT-5.6 Sol Ultrafast hits 750 tokens a second, with no price by stephane 3 September 2026 0 Cerebras published a bar chart with a 14x on it, and the bar beside it says 2.5x. That is the... Read moreDetails
Uber and Pony.ai promised 2,000 robotaxis and named one city by stephane 3 September 2026 0 Ten cars. That is the whole robotaxi fleet carrying paying passengers in Europe on this partnership right now, a handful... Read moreDetails
Gemini 3.7 Flash is half price until 31 December by stephane 3 September 2026 0 We had the Gemini pricing page open for something else, and the 3.7 Flash row carried a sentence most of... Read moreDetails
Grok 4.6 keeps $2/$6 and adds a 200K pricing cliff by stephane 3 September 2026 0 The headline on xAI's page is the price that didn't move. Grok 4.6 landed on 12 August at the same... Read moreDetails
Twitch’s AI training toggle is on, and it covers your chat by stephane 13 August 2026 0 You went to bed a streamer and woke up a training set. On 12 August Twitch shipped a setting called... Read moreDetails
DeepSeek V4 Pro 0813 is live, at 3x the Flash price by stephane 12 August 2026 0 Nothing in your code has to change, again. On 12 August DeepSeek pointed deepseek-v4-pro at a new build called DeepSeek-V4-Pro-0813,... Read moreDetails
Nemotron 3.5 Lightning runs 30B on 3B active parameters by stephane 3 September 2026 0 Thirty billion parameters, and only three billion of them wake up per token. That ratio is the whole pitch for... Read moreDetails