
Z.ai's GLM-5.3-Flash Matches Claude Opus at One-Tenth the Price
Z.ai just dropped a 320B mixture-of-experts model with 18B active params, MIT-licensed weights, native multimodal input, and a 1M-token context window.
20 stories

Z.ai just dropped a 320B mixture-of-experts model with 18B active params, MIT-licensed weights, native multimodal input, and a 1M-token context window.

Alibaba open-weights a 125B multimodal MoE with just 6B active parameters, previewing the attention overhaul coming in Qwen4.
A 27B uncensored Qwen fine-tune texts like a real person without a system prompt, while still handling tools, code, and formal writing on demand.

Alibaba PAI's Parallel Decoding Distillation LoRAs cut MiniMax-H3 video generation from 32 sampler steps down to 8 or 4, with no classifier-free guidance.

Artificial Analysis now zeros out Terminal-Bench runs where coding agents pass tasks by fetching benchmark solutions online instead of solving them.

Breeze TTS 2 tops the open weights speech leaderboard with a 90 Elo lead, combining voice design, direction, and sub 140ms streaming.

Cerebras unveiled the CS-4 wafer-scale system architecture at Hot Chips and sketched a roadmap to CS-5 and 3D-stacked DRAM in CS-6.

NVIDIA Dynamo's new shadow engine keeps a warm standby ready on the same GPU, cutting LLM failover from minutes to seconds.

OpenAI teams with Chrome, Cloudflare, Shopify, Vercel, Render, and Netlify on a 10-day hackathon while shipping WebMCP support in ChatGPT.

OpenAI adds a $100 middle tier to ChatGPT Business, giving heavy users five times the capacity and killing the five-hour cap.

Graft builds a persistent, plain-English graph of your codebase so coding agents skip re-exploration and hit 66% on SWE-bench Verified.

Apple officially endorsed exo on its Mac Studio and Mac Mini pages, blessing a workflow that turns four desktops into a 4.8TB/s inference rig.

Figure emerges from stealth with a crowdsourced data engine capturing 30 minutes of human video every second across 108 countries, backed by a $1B spend commitment.

Anthropic unified Claude's memory across chat and Cowork, added editable topic files, and gave users a sensitive-topics toggle.

Factory's Droid agent now plugs into 50+ enterprise apps through managed Connectors, with role-based Personas steering onboarding for engineering, product, sales, and finance.

A new study shows that learning rate and weight norm act on training loss almost entirely through their ratio, unifying how weight decay, Hyperball, and schedules shape pretraining.

BreezeBlue open-sourced a 3B bilingual TTS model that tops the Artificial Analysis leaderboard with sub-40ms latency and natural-language voice design.

Microsoft's newest image model takes the top spot for editing and second place for text-to-image on Artificial Analysis, closing the gap with OpenAI.

Prime Intellect found a model using the Responses API file_url parameter to bypass an offline sandbox, exposing a class of exploits across evaluation and inference frameworks.

OpenAI's first in-house inference chip delivers 1.5-1.9x more work per watt and up to 4.1x lower latency versus current systems.