Mistral AI releases Timestral forecasting models
MarkTechPost: Mistral AI Releases Timestral: A 209M Forecasting Model Family That Outperforms Google’s TimesFM-3
Mistral AI has released Timestral, its first family of time series foundation models. Timestral is a set of 4 zero-shot forecasting models from Mistral AI, sized from 5M to 209M parameters. The largest checkpoint ranks first among pretrained, zero-shot models on GIFT-Eval, TIME and fev-bench. It does this without a GPU at inference. For developers, this Timestral time series foundation model means probabilistic forecasts on standard server CPUs, with open weights and inference code. TL;DR
Get smart on it
Timestral is a family of time series forecasting models released by Mistral AI, available in four sizes ranging from 5 million to 209 million parameters, designed to predict future values in sequential data without requiring specialized training on new datasets. The largest model ranks first among comparable zero-shot forecasting models on three major benchmarks and can run on standard server CPUs without a graphics processor, making probabilistic forecasting more accessible to developers. Time series forecasting typically requires either fitting separate classical models to each dataset or retraining deep learning models from scratch, whereas Timestral transfers learned patterns across different data domains. The model uses a causal patch-based transformer architecture and outputs 21 quantiles per prediction step, allowing it to handle multiple related variables and different types of input information simultaneously.
Relevant courses
Want to go deeper than the news? Explore live, cohort-based AI courses taught by practitioners.
Browse AI courses on MavenRelated news
Nvidia launches new open models to run agents in the enterprise and on PCs
Nvidia is leaning hard into the trend of enterprise companies wanting to run their own open models in their own data centers and also trying to make on-device, private AI agents a thing.
Welcome Muse Spark Open: open-weight frontier multimodal mixture-of-experts model from Meta
Open-weight models let researchers and companies build AI systems without relying on proprietary providers, potentially accelerating innovation across industries.
Also reported by The Verge, The Decoder, TechCrunch and SiliconANGLE
MiniMax Just Launched M3.1 Flash: 428B MoE Agentic Model at $0.15/M Input
MiniMax has just released MiniMax M3.1 Flash on its API, priced at $0.15 per 1M input tokens. MiniMax also confirmed that Space Bunny was an earlier M3.1 Flash build. That stealth model processed over 200T tokens. For developers, the message is near-frontier agentic scores at Flash-tier cost. Model weights are due in late October, 2026. TL;DR Size: 428B total parameters, about 23B active (MoE); 1,000,000-token context Runs on: MiniMax API, M Plan and MiniMax Code today. Self-hosting h
OpenAI will reportedly ship at least one model this year but its flagship still needs alignment work
OpenAI is racing to finish its next flagship model, which uses the same looped transformer technique as Astra to boost math and coding performance. The tradeoff is that looping makes safety monitoring harder. Microsoft accidentally revealed details about how the technique works on a public page. Whether or not the flagship is ready in time, employees say one or two new models will likely ship by year's end. The article OpenAI will reportedly ship at least one model this year but its fla
Open From Data Center to Desktop: NVIDIA Expands Nemotron Family With Highest Efficiency Super VL and Bolt Models
NVIDIA today released Nemotron 3.5 Super VL, an open 120-billion-parameter model that delivers the fastest task completion and highest accuracy for specialized always-on agents. Joining it are two new edge models: Nemotron 3.5 Bolt 74B-A9B 24GB, optimized to balance intelligence and speed per GB for local agentic workloads, and Nemotron 3.5 Bolt E4B 4GB, optimized for intelligence per GB for fast conversational agents. Super VL and Bolt 74B-A9B are available now, with Bolt E4B coming Oct. 23. T
Also reported by The Decoder, SiliconANGLE and MarkTechPost
Meet Kyutai REKI: The 2.4B VLA Model Bringing Efficient Memory to Robot Manipulation
Kyutai, with Inria and École des Ponts, has released REKI, a 2.4B vision-language-action model (VLA) for robot manipulation. The REKI vision-language-action model is pretrained only on public data. It also carries a cheap memory of its recent past. On three out-of-distribution benchmarks, only models about twice its size beat it. All weights are live on Hugging Face today. TL;DR Size: 2.4B parameters (2.2B Qwen3.5-2B VLM + 0.2B flow-matching action expert). Memory covers the last 3 fram