In productionAI
MiniMax H3 opens its weights for unified video-audio workflows
Open weights expose two checkpoints: FL2VA for text-to-audio-video and Ref2VA for multimodal references.
Written by AI1 source2 min
Loading…
Models, agents, applications: artificial intelligence put to work on real problems.
Open weights expose two checkpoints: FL2VA for text-to-audio-video and Ref2VA for multimodal references.
Readers can mark a website as a preferred source for Google directly on the publisher’s site.
Pocket turns text prompts into small interactive games that react to touch and phone movement.
University at Buffalo’s CrowdHydrology network spans 26 states and has collected more than 20,000 readings.
A team from RAI Institute and Boston Dynamics is training robots using motion data, videos or animations.
Slack Code creates a task-specific channel when a team mentions an AI coding agent.
University students outside the US, including those in Japan, are eligible; the application deadline is December 31, 2026.
Version v0.1.0-rc.8 adds native image requests and mixed text-image tasks.
The glasses start at 2,699 yuan and target people who wear glasses for more than 16 hours daily.
The beta uses Meta’s Muse Spark model to answer questions from a shared screen.
Hierarchical CoAtNet 1 recognized yoga poses with more than 93% accuracy during testing.
The Southampton team found excess and enlarged centrosomes are separate defects that can occupy different tumor regions.
It improved accuracy by 7.9% over its proprietary base model and 9.8% over evaluated open-source models.
IPO-Mine splits filings into sections and extracts charts for machine analysis.
Alipay has released the AHA multi-agent interoperability protocol, enabling agents on different devices to discover one another and pass along tasks.
The model combines 2.8 trillion parameters with a 1-million-token context window.
The opt-in beta can find previously viewed pages, preview them and group duplicate tabs.
The custom harness raised Claude Opus 5 from 30% to 100% on an interactive reasoning benchmark.