Kimi K3.1 Pro launches with API rates one-third below K3
The model accepts text, images and video and supports JSON-schema responses.
Models & products
New AI model releases, updates and pricing changes, from primary sources.
Give a rough 3D mesh more detail: Hi3D releases refinement weights
The Hi3D team released v1.0 model weights and inference code on October 23.
GLM Flash comes to Mistral at $0.15 per million input tokens
Mistral added Z.ai’s GLM-5.3-Flash to its API offerings on October 22.
Same GPT-4o name, different shutdown dates: which version does your app use?
October 23, 2026 is the scheduled shutdown date for a group of older OpenAI API models.
Plan first, render second: Canva’s new image model Craft
Canva announced Craft, its own image model that plans composition before rendering.
Protein AI that designs and searches: NVIDIA and UW unveil RFD4-Proteina
NVIDIA and UW's Institute for Protein Design announced an open model combining design with search.
New US lab Reflection releases Beam, a 501B-parameter open model
Reflection released its first model, Beam, with open weights and deployment and post-training materials on Hugging Face.
NVIDIA releases open Nemotron 3.5 models for everything from servers to laptops
NVIDIA released Nemotron 3.5 Super VL, an open model for agents, plus two Bolt models for PCs and edge devices.
FLUX.3 Image Klein arrives as an open image model with bounding-box layout control
Black Forest Labs' FLUX.3 Image Klein image model is out and works in Hugging Face Diffusers from day one.
Google launches Gemini 4 Argon, a frontier model aimed at coding and professional work
Google unveiled Gemini 4 Argon, its latest frontier model, on October 13 and released it globally the same day.
Meta unveils Muse Wonder, agent models built to remember and follow through
Meta unveiled Muse Wonder and Muse Wonder Full on October 12, calling them its most capable models to date.
Google's EmbeddingGemma 2 brings on-device search across text, images, audio and video
Google DeepMind released EmbeddingGemma 2, a lightweight embedding model that maps text, images, audio and video into one space.
