Reanimate Gives Claude a Ruler to Clone Web Animations Within 0.03% Pixel Error
A new Claude Code skill records animations from any site, video, or gif, then rebuilds and verifies the copy frame by frame until the motion matches.
20 stories
A new Claude Code skill records animations from any site, video, or gif, then rebuilds and verifies the copy frame by frame until the motion matches.
Exo lets an agent inspect and rewrite its own harness code at runtime, cutting model costs by double digits while matching task completion.
A new flow-matching method enforces Bellman consistency at every noise level, producing accurate return distributions and stronger offline RL from limited data.
Liquid AI's LFM3 family packs text, vision, audio, omni, and decision models into six 2B-class checkpoints built from one tri-hybrid backbone.
Ollama picks the window, flash attention, and an MTP draft before you type. Here is the one number that judges each, on a 24GB card and a Mac.
Google DeepMind's Gemini-powered PlanAssist cut householder planning assessment times by 58% in UK pilots and now expands to 12 more councils.
Sperid Labs released Idol, an Apache 2.0 image-to-3D model that generates up to 1M Gaussian splats in seconds on consumer GPUs.
Pika Labs just dropped open LoRA adapters on top of MiniMax-H3 that generate video and synchronized audio in only six sampling steps.
DeepSeek quietly open sourced the research prototype behind its internal coding super-agent, a repository-scoped assistant built to minimize human attention.
Agent Sprite Forge 0.4 turns one approved character still into a full game-ready sprite set with animations, loops, and engine exports.
Mika is an open-source Android agent that drives any app on a hidden background screen, browses the web, and runs scripts in a local Linux sandbox.
A 9B reasoning model distilled from DeepSeek-V4-Pro and decensored with Heretic now ships as imatrix GGUFs that run on 4GB GPUs.
Mainplane is an open-source agent harness that replaces lossy context compaction with full-session retrieval, cutting cost and latency on long-running agent runs.
Anthropic set thousands of Claude agents loose to generate training data that slips past safety monitors but still produces misaligned models.
A new compute-in-memory accelerator for small language models hits 400 tokens/s on LLaMA3.2-1B with 49x better energy efficiency than Jetson Orin Nano.
Artificial Analysis refreshed its knowledge and hallucination benchmark, cutting it to 3,000 questions and exposing which models were quietly overfitting.
A crowd-sourced ternary quantization of Qwen3.8-27B runs 27B parameters at 25 tokens per second on a base M5 MacBook in 13 GB of RAM.
Aura open-sourced Autopilot, an in-editor agent that playtests Unreal, Unity and Godot games, beating Claude Code on accuracy and speed in a seven-task benchmark.
NVIDIA's new 9.6M parameter RE-USE model strips noise, echo and distortions from speech across 7 sampling rates and beats models 25x larger.
Gemini 4 Carbon reshapes the Pareto frontier on three leaderboards, GPT-6.1 Terra undercuts rivals on cost, and Agent Mode gains tool connectors.