New releases

Mistral · Mistral's new speech-to-text and text-to-speech models, with a voice agent SDK built on LiveKit and Pro Voice cloning from up to 30 minutes of audio

SenseTime · SenseTime's 27B embodied model for robot planning and decision-making, trained from Qwen3.8-27B with experience from its open-source SenseNova-RoboRSI harness…

Bindwell · Bindwell's foundation model for predicting experimental outcomes in molecular discovery, trained across 800M–9B sizes on about 80M experimental measurements

Cloudflare · Cloudflare's System One extraction model, which extracts, computes, estimates and forecasts values from unstructured inputs

Dream Labs · Dream Labs' World Action Model, a robot foundation model that the lab says controls unseen bimanual parallel-gripper robots zero-shot on unseen tasks

Nace.AI · Nace AI's multimodal decision model for text, images and video, available through Nace at $0.06 per 1M input tokens

fal · A post-trained image-to-video model from fal Research, built in partnership with Black Forest Labs and available on fal

GoTo (with AI Singapore) · GoTo's 27B Gemma 3 instruct model for English and Indonesian regional languages (Javanese, Sundanese, Batak Toba, Balinese), developed with AI Singapore

Inception · Inception's diffusion LLM with a 1M-token context, live on the Inception API, OpenRouter, Baseten and Vercel

Ant Ling (InclusionAI, Ant Group) · InclusionAI's open Ming-3.0 audio family: multilingual TTS with voice cloning, Chinese/English ASR with dialects, and native audio-language MoE models for…

Dream · Dream's open cybersecurity language models (8B and 30B-A3B), with a dedicated security mid-training stage before chat tuning

Mind Lab (Macaron AI) · Mind Lab's System 1 spatial decision model on Qwen3.5-4B: scores answer options, value bins and image regions with native attention Q/K, without autoregressive…

JPMorgan Chase · A time-series forecasting foundation model from a JPMorgan Chase team, announced in a thread by a team member

Ropedia · Ropedia's egocentric world foundation model trained on about 10k hours of first-person video for loco-manipulation that transfers across robot embodiments

Skyfall AI · Skyfall AI's Engineering World Model for CAD and BIM tasks

Abacus.AI · Abacus.AI's open-weights detector for AI-generated images, released on Hugging Face under Apache 2.0 with GGUF builds

Viggle AI · Viggle's physics-native 3D world foundation model: scenes and characters with motion, built in 3D rather than pixel/video space

VAST · VAST's real-time world-model renderer that turns a rough game or reference video into detailed frames while upstream code keeps the world state

Blockway · Blockway's open-weights 32B model for agentic coding, the full release after its Preview, tuned for long agent sessions

OrcaRouter · A 27B open-weight vision-capable model built for defensive cybersecurity work such as vulnerability research and incident response

Birla AI Labs (Aditya Birla Group) · Birla AI Labs' first time-series foundation model: a 271.8M-parameter model for zero-shot probabilistic forecasting that returns a range of likely values…

Kyutai · Kyutai's 2.4B vision-language-action robot model, built with Inria and École des Ponts, pretrained on public data only, with a built-in memory of its recent…

NVIDIA · NVIDIA's 4B looped model: the same weights can run 1, 2 or 4 loops per token, trading compute for accuracy

Knowledgator · Knowledgator's family of instruction-following classifier encoders, from a 435M Edge model up to a 27B model