
LM Arena's AutoEval Ranks AI Models in Hours, Beating Frontier LLM Judges
Arena launches AutoEval, a reward-model-based system that scores new AI models in hours instead of days, with >0.98 correlation to live human votes
20 stories

Arena launches AutoEval, a reward-model-based system that scores new AI models in hours instead of days, with >0.98 correlation to live human votes

OpenAI slashes GPT-5.6 Luna prices by 80% and Terra by 20%, adds 2.5x faster Sol mode just three weeks after launch

Simile raises $200M at a $2B valuation to build a foundation model that simulates all 8 billion humans, with Fortune 100 customers already running tens of millions of predictions

Google and RadixArk bring SGLang's full inference stack to TPUs, letting teams swap GPU clusters for cheaper hardware without rewriting a line of serving code

Loka post-trained Arcee's open 26B Trinity Mini with RL to run multi-step biomedical research workflows, hitting 81.2% on drug-tool evaluation and 0.863 on biological reasoning

Perplexity's Spaces become Projects: a persistent file system and self-improving Brain memory now power long-running agentic work for all users

Arena.ai adds seven granular categories to its image editing leaderboards, revealing GPT Image 2 leads every dimension across 28M+ human votes

Goodfire's Silico platform exposes competing neuron groups inside LLMs and lets you surgically boost safety signals without retraining the model

Google DeepMind launches Gemini Robotics 2, bringing full-body humanoid control, multi-finger dexterity, and cross-robot collaboration to physical AI
Sarvam's new Saaras V4 Multi-speaker handles overlapping conversations in a single pass, while claiming state-of-the-art English ASR across seven global benchmarks.

Luma Agents adds Layers: extract, isolate, and re-edit any element in an image through natural conversation, powered by Uni-1

Unsloth shipped Dynamic 2.0 GGUF quants of Thinking Machines' 276B Inkling-Small, letting the multimodal MoE run locally in llama.cpp, Ollama and LM Studio.

CodePilot is a free, open-source desktop AI agent that connects 17+ providers, adds MCP tools, multi-agent orchestration, and phone-based remote control — all in one app.
Sarvam AI's Bulbul V4 upgrades its Indic TTS engine with richer emotion, natural expression, and a wider vocal range across 11 Indian languages.

Agnes AI's new reasoning model scores 39 on the intelligence index with top-tier coding performance, at just $0.45/$0.90 per 1M tokens

A community repackage turns Anthropic's Claude Design workspace into a portable Agent Skill that runs inside Cursor, Claude Code, and Codex.

ThunderAgent fixes KV cache thrashing in agentic inference, delivering 2.5x higher throughput and ~10x lower latency -- accepted as an ICML 2026 Spotlight paper.

MLX-Audio brings a full on-device audio stack to Apple Silicon, covering TTS, STT, and speech enhancement with 30+ models and no cloud dependency.

SpaceXAI's new voice model hits #2 on the Speech-to-Speech Index with 0.70s latency, beating GPT-Realtime at half the price

LMSYS brings end-to-end MXFP8 and NVFP4 RL training to Miles, matching BF16 reward curves while cutting rollout time on Blackwell B200 GPUs.