PyTorch 2.14 Ships Faster Apple Silicon, Rebuilt Distributed Training and NVGEMM
The latest PyTorch release ships NVGEMM CUTLASS kernels for Inductor, a rebuilt nccl2 backend, native Apple Silicon linear algebra, and first-class fault tolerance.
// Intelligence
Today in AI
A focused feed of models, agents, research and open-source releases for people building with AI.
Updated
Nodeterm's Open-Source Canvas Lets Developers Run Claude and Codex Side by Side
Nodeterm turns terminals and coding agents into draggable nodes on an infinite canvas, with tmux-backed sessions that survive reboots.
Nous Research's Hermes Agent Pantheon Connects All Your AI Agents Into One Interface
Hermes Agent v0.21.0 Pantheon lets the desktop app hold live connections to multiple gateways at once, unifying every bot and chat under one roof.
Lovable Now Builds Full-Stack Apps Directly Inside Slack
Lovable now lives inside Slack as a chat agent, letting teams spin up projects, ship updates, and query app data without opening a browser tab.
UC Berkeley Team Cuts Diffusion Model Training Time by 40% With Frequency-Aware Loss
A new frequency-domain training objective for pixel-space flow matching cuts convergence time by up to 40% without touching the architecture.
Treg Gives AI Agents Access to 2,600 Paid API Tools for Pennies
Treg is an open-source proxy that gives agents access to 2,600+ paid APIs through one token, billed per call with no vendor signups.
Google DeepMind's Gemini 3.8 Flash Beats Claude Opus 5 at 6x Lower Cost
Google's third Flash release in as many months pushes a workhorse model into frontier-tier territory on agentic coding, legal, and finance benchmarks.
Rebuilding the Runtime: What NVIDIA NOOA Reveals About Agent Frameworks
Its object-oriented model clarifies the trade-off between live program state and durable runtime control.
Runway Dev MCP Lets AI Coding Agents Control Generative Media APIs Directly
Runway's new hosted MCP server plugs its generative media API directly into coding agents like Cursor, Claude Code, and Codex for research, routing, and debugging.
Multiverse Computing's Quasar 438B Tops Europe's AI Leaderboard While Beating Bigger Models
Multiverse Computing's Quasar 438B tops European AI rankings with a 43 Intelligence Index score and 15.3-second reasoning responses.
Amap's ABot-Recon Rebuilds 3D Worlds From Live Video Using Under 7GB
Amap CV Lab open-sourced ABot-Recon, a streaming 3D reconstruction model that runs on a rolling 12-frame window at 24 FPS.
BytePioneer-AI's Codex-Host Lets Developers Run Claude Code and Pi in One Window
BytePioneer-AI released codex-host, an open-source layer that runs Pi and Claude Code as native sessions inside Codex Desktop.
Video-talkcraft Turns Claude Code Into a Frame-Perfect Video Director
A new open-source agent skill hands Claude Code and Codex a full motion-design studio for voiceover explainer videos with word-level sync and Remotion rendering.
Utopia Gives Open-Source RAG a Memory of What Was True When
Utopia is a self-hosted Rust and Postgres knowledge platform that layers RAG on top of a bitemporal knowledge graph tracking when each fact was true.
Alibaba's Qwen3.8-Max-0902 Tops Code Arena, Beating Claude Opus 5
Alibaba's flagship 2.4T-parameter model gets a coding and agent-work refresh with the same pricing and 1M context window.
alphaXiv Opens OpenResearch, Turning Coding Agents Into Autonomous Scientists
alphaXiv released a local-first, MIT-licensed CLI that spins up parallel Claude Code, Codex, or OpenCode agents against isolated git worktrees.
Anthropic's Claude Fable 5.1 Tops Every Frontier Model but Costs 20% More
Independent evaluations put Anthropic's new flagship at the top of the intelligence charts, but the win comes with a token-count tax that eats the cache savings.
audio.cpp Runs Speech and Music AI up to 5x Faster Without Python
A pure C++ inference engine built on ggml runs TTS, ASR, voice cloning, and music generation up to 5x faster than Python baselines with no Python dependency.
MIT's SwarmWorld Shows 200 AI Agents Building a Society Without Talking
MIT researchers put hundreds of identical LLM agents into a persistent world and watched them evolve specialization, tool inheritance, and technology that survives without them.
Tsinghua's Puro-2B Trains a Language Model for Less Than One GPU Rental
A Tsinghua team trained a 2B language model from scratch on consumer RTX 5090s for under $6,900, matching Qwen2.5-1.5B and open-sourcing everything.
Microsoft Finds a 2023 Trick Beats Two Years of Linear Attention Research
A new benchmark study finds that sliding-window attention with sinks matches or beats post-trained linear attention models, without any retraining.
Cursor's Claude Fable 5.1 Hits 73.4% and Catches Its Own Bugs
Cursor added Anthropic's new Fable 5.1 model, which posted a 73.4% score on CursorBench 3.2 and excels at self-verifying long coding runs.
Lovable's Fable 5.1 Fixes Live Apps 17% Better at 31% Lower Cost
Lovable's new default coding model gains 17% on the hardest fix-and-iterate tasks while cutting cost by 31%, with sharper self-verification and visual taste.
Runway's Ruby Now Exports AI Footage Directly Into Pro VFX Pipelines
Runway Ruby now exports scene-referred half-float EXR sequences in ACEScg 1.3 and 2.0, slotting AI video directly into professional color pipelines.
No stories match these filters.