Claude Code Turns Any Topic Into Animated Explainer Videos Instantly
A Claude Code skill reverse engineers viral Douyin knowledge shorts, generating storyboards and rendering finished videos frame by frame with Canvas and Playwright.
20 stories
A Claude Code skill reverse engineers viral Douyin knowledge shorts, generating storyboards and rendering finished videos frame by frame with Canvas and Playwright.
Artificial Analysis released AA-3D, a benchmark that ranks 21 AI models by human-judged quality of interactive three.js scenes built from written briefs.
Ai2's TutorMoments benchmark replays real math sessions to test whether LLMs scaffold learning or spoil it by over-helping students.
Speridlabs released a 3B pixel-space diffusion transformer that skips the VAE entirely, generating 1024px images and doubling as a depth and restoration backbone.
Factory AI makes Environments generally available, giving its Droid coding agents pre-built, continuously refreshed sandboxes to parallelize work across fleets.
A training-free attention trick lets diffusion transformers like FLUX.1 render up to 6x their training resolution, while running up to 3.2x faster.
Artificial Analysis launched a leaderboard for Cognition's FrontierCode 1.1, a benchmark that grades whether a model's code would actually get merged.
Meta unveils Muse Wonder and Muse Wonder Full, frontier models built for long-running agent work with Active Memory and Active Focus.
Mistral's trillion-parameter Le Chonk model hits 42% on Dense 200 visual grounding, edging past frontier closed models and targeting industrial inspection work.
A new open-source desktop app bundles IndexTTS, VoxCPM2, OmniVoice, and Qwen3-TTS for local voice cloning on Mac and Windows, with no cloud fees.
Microsoft Research built personalized AI student simulators that mimic real learners, giving tutor models fast feedback without costly human studies.
Z.ai has released GLM-Legal, a specialized model built on GLM-5.3-Flash that targets contract review, evidence analysis, and compliance workflows.
UkisAI released compact 2-bit and 3-bit GGUF quantizations of Swift 1.5 that run a 27B reasoning model in under 10GB while cutting thinking tokens by more than half.
NVIDIA's new open 120B vision-language MoE activates just 12B parameters per token, handles 1M context, and runs on a single DGX Spark at NVFP4.
A new open-source pipeline takes a single-person video and spits out a Mixamo-rigged GLB you can drop straight into Blender or Unity.
Alibaba's Qwen team unveils an elastic RL framework that trains agents on hours-long tasks, cutting training time by up to 44% while lifting repo-generation accuracy by 6 points.
A full set of imatrix GGUF quantizations lets you run Qwen2.5-32B-Instruct locally, from a 9GB phone-friendly build up to lossless 65GB weights.
ChatR Labs released a rank-128 LoRA that handles both text-to-image and reference-based image editing on Qwen-Image-2.1 using one adapter.
Project Maya runs the 321B-parameter GLM-5.3-Flash on a single or dual NVIDIA GPU rig by streaming experts from VRAM, RAM, and NVMe.
A new open-source tool captures your mobile app traffic, then has Claude generate a typed Python client so you can script any app like a library.