FundingTopic: Generative Image & Video
Crest summary (AI-written):Utopai, an AI film studio, raised $60M at a $1B valuation, according to The Information.
Companies: Utopai
OtherTopic: Generative Image & Video
Crest summary (AI-written):Fable 5.1 hides messages in images, according to a Hacker News discussion of the technique.
EngineeringPrimary topic: Generative Image & VideoAlso: Cloud Infrastructure
Crest summary (AI-written):HeyGen ported its 18B+ parameter Avatar IV video generation model to Google Cloud's Trillium (v6e) TPUs using torchax and XLA.
Companies: Google CloudHeyGen
EngineeringPrimary topic: Generative Image & VideoAlso: ML Frameworks & Training
Crest summary (AI-written):Google Developers Blog describes implementing Sparse VideoGen to address the quadratic latency bottleneck of self-attention in high-resolution video diffusion models on TPUs.
Companies: Google
LeadershipTopic: Generative Image & Video
Crest summary (AI-written):Sean Parker is rebuilding Stability AI around music, according to TechCrunch.
Companies: Stability AI
LaunchTopic: Generative Image & Video
Crest summary (AI-written):An open-source Lego AI generator was shared on Hacker News.
LaunchTopic: Generative Image & Video
Crest summary (AI-written):Tavus unveiled Griffin, which it calls the first Human Interaction Model, saying it passed a video Turing test in which 48% of users thought it was human during live chats.
Companies: Tavus
LaunchTopic: Generative Image & Video
Crest summary (AI-written):A developer demonstrated giving the model Opus 5.5 a simulated paint canvas, sharing the project on Hacker News.
LaunchTopic: Generative Image & Video
ResearchPrimary topic: Computer VisionAlso: Generative Image & Video
Crest summary (AI-written):A new AI tool can reconstruct what a person is looking at from brain scans, and also predict brain activity from viewed images, according to MIT Technology Review.
LaunchTopic: Generative Image & Video
Crest summary (AI-written):Google DeepMind introduced Gemini 3.8 Live with Live Avatar on 2026-09-24.
Companies: Google DeepMind
OtherTopic: Generative Image & Video
LaunchTopic: Generative Image & Video
Crest summary (AI-written):Higgsfield AI shipped new video features in a single day using OpenAI's GPT-6 Astra, according to OpenAI's announcement.
Companies: Higgsfield AIOpenAI
EngineeringPrimary topic: Generative Image & VideoAlso: Developer Tools
Crest summary (AI-written):Hugging Face published a rebuild of the AUTOMATIC1111 interface using a Gradio workflow on 2026-09-10.
Companies: Hugging Face
LaunchTopic: Generative Image & Video
Crest summary (AI-written):OpenAI introduced ChatGPT Images 2.5 on 2026-09-08, an updated image generation model for ChatGPT.
Companies: OpenAI
LaunchPrimary topic: Generative Image & VideoAlso: API & Integration
Crest summary (AI-written):OpenAI's GPT Image 2.5 Flare and GPT Image 2.5 Sunburst are now available on Vercel's AI Gateway. Both models accept text prompts and reference images.
LaunchPrimary topic: AI AgentsAlso: Generative Image & Video
Crest summary (AI-written):Google DeepMind introduced Gemini with agentic video understanding, a capability that lets the model reason over video content. The announcement was made on 2026-09-01.
Companies: Google DeepMind
LaunchPrimary topic: Generative Image & VideoAlso: LLM Infrastructure
Crest summary (AI-written):Meta Superintelligence Labs' first image model, Muse Image, is now available on Vercel's AI Gateway. It is a separate family from Muse Spark and returns images rather than text.
Companies: Meta Superintelligence LabsVercel
UpdatePrimary topic: Generative Image & VideoAlso: API & Integration
Crest summary (AI-written):Vercel's AI Gateway now supports asynchronous video generation, so generateVideo no longer has to hold a single HTTP request open until the result is ready.
Companies: Vercel
LaunchPrimary topic: Generative Image & VideoAlso: API & Integration
Crest summary (AI-written):Alibaba's Wan 3.0 video generation model is now offered through Vercel's AI Gateway under the identifier alibaba/wan-v3.0-video. The model combines text-to-video, image-to-video, first- and last-frame conditioning, and reference-based generation in a single model.
LaunchTopic: Generative Image & Video
Crest summary (AI-written):xAI's Grok Imagine Image 2.0 Preview is now available on Vercel AI Gateway, where it follows detailed instructions and plans typography and layout together.
LaunchTopic: Generative Image & Video
Crest summary (AI-written):ByteDance's Seedance 2.5 is now available on Vercel AI Gateway, generating clips up to 30 seconds with synchronized audio and accepting up to 50 reference inputs per request.
LaunchTopic: Generative Image & Video
Crest summary (AI-written):MiniMax H3 is now available on Vercel's AI Gateway, generating 2K video from text prompts, a starting image, first and last frame pairs, or reference material.
LaunchPrimary topic: Generative Image & VideoAlso: Audio, Video & Images
Crest summary (AI-written):Google DeepMind launched Lyria 3.5 in Google Flow Music, claiming improvements in musicality, lyrics, vocals, and creative control.
Companies: Google DeepMindGoogle Flow Music
LaunchPrimary topic: RoboticsAlso: Generative Image & Video
Crest summary (AI-written):NVIDIA released Cosmos-H-Dreams, a real-time generative simulation model for surgical robotics, announced by Hugging Face on 2026-07-27.
Companies: Hugging FaceNvidia
