GPT-56T 861 —
MUSE-SPK 835 -0.7%
GPT-56SC 828 -5.2%
QWEN-38X 824 —
CL-OP55X 822 —
GROK-46H 822 -5%
GPT-6A 820 —
GLM-5 784 -8.4%
CL-FAB5H 743 -5.6%
KIMI-K3X 742 -8.4%
CL-OP5H 720 -5.8%
CL-OP5X 709 -18%
CL-OP46H 698 -5.9%
CL-OP47H 690 -5.9%
GEM-38FH 677 +0.1%
GEM-37FH 657 -24%
GPT-56S 622 —
CL-OP47 582 -0.7%
GPT-55H 582 —
INKL 531 —
GEM-31P 513 —
GEM-3P 499 —
CL-OP46 496 -0.2%
CL-OP48 490 —
GPT-56T 861 —
MUSE-SPK 835 -0.7%
GPT-56SC 828 -5.2%
QWEN-38X 824 —
CL-OP55X 822 —
GROK-46H 822 -5%
GPT-6A 820 —
GLM-5 784 -8.4%
CL-FAB5H 743 -5.6%
KIMI-K3X 742 -8.4%
CL-OP5H 720 -5.8%
CL-OP5X 709 -18%
CL-OP46H 698 -5.9%
CL-OP47H 690 -5.9%
GEM-38FH 677 +0.1%
GEM-37FH 657 -24%
GPT-56S 622 —
CL-OP47 582 -0.7%
GPT-55H 582 —
INKL 531 —
GEM-31P 513 —
GEM-3P 499 —
CL-OP46 496 -0.2%
CL-OP48 490 —

Live Feed

6d ago model

Cactus Needle 3: An 8-29MB Model That Matches DeepSeek V4 Flash on Tool Calls After One Fine-Tuning Epoch

Cactus Compute's Needle 3 is a 29-121M parameter foundation model that fits in a single 8-29 MB binary and runs at 400-4,000 tokens per second on a Raspberry Pi. Its 4-layer subnetwork matches DeepSeek V4 Flash on targeted automation tasks after a single fine-tuning epoch.

6d ago research

Anthropic's New R&D Index: Claude Leads 26% of Internal AI Work, 90%+ at AI-Collaborates Level

Anthropic published its first R&D Automation Index today, cataloguing every AI R&D task at the company and rating its automation level. As of August 2026, Claude leads 26% of Anthropic's AI R&D end-to-end, and more than 90% of work sits at or above the AI-collaborates threshold.

6d ago policy

US Military Almost Boarded a Chinese Ship After AI Hallucinated a Nuclear Threat

A special operations analyst used an AI chatbot to assemble an intelligence report during the Iran war this spring. The bot invented a nuclear weapons shipment on a Chinese vessel. Armed personnel were airborne and preparing to board before the error was caught.

6d ago release

Huawei Atlas 960E SuperPoD: 8 EFLOPS at FP8 Across 4096 NPUs, Replacing 48,000 Optical Modules With 5,500

Huawei launched the Atlas 960E SuperPoD at HUAWEI CONNECT 2026 in Shanghai, claiming the industry's first NPO-based SuperPoD architecture. The system delivers 8 EFLOPS at FP8 precision across 4096 NPUs, cuts power by 550 kW versus conventional interconnect, and targets 10-trillion-parameter model training.

6d ago policy

Unsealed NYT Filings: Microsoft Director Called AI Training Scraping the Largest Theft of Labor in Human History

New unredacted documents in the New York Times copyright lawsuit against OpenAI and Microsoft reveal a January 2023 internal memo from Microsoft director of applied science Brent Hecht describing the practice as an astonishing theft of unprecedented proportions. Copilot reduced NYT click-through rates by up to 93 percent.

6d ago policy

ZCode Silently Packages Entire Git Histories and Uploads Them to Alibaba Cloud — No UI Toggle Stops It

Reverse engineering by developer ferstar reveals Z.ai's ZCode coding app captures full workspace snapshots including complete .git histories and uploads them to Aliyun OSS. The encryption key is held only by Z.ai's servers. Privacy toggles do not disable the behavior.

7d ago release

Devin's Code Scans: Agentic MapReduce Turns Engineering Goals into Merged PRs at 96%

Cognition launches Code Scans in Devin, using Agentic MapReduce to fan out codebase investigations across parallel agents and synthesise findings into ready-to-review pull requests. Philips reports a 96% PR merge rate and over 700 engineering hours saved in early testing.

7d ago release

Qwen3.8-Omni-Flash: 1M Context, Four Modalities, Audio Surpasses Gemini 3.8 Flash

Alibaba releases Qwen3.8-Omni-Flash, a native omnimodal model accepting text, image, audio, and video in a 1M-token context. Overall audio performance exceeds Gemini 3.8 Flash; API price per audio hour drops 98% from the prior generation.

7d ago funding

Crusoe Raises $3.9B at $30.9B Valuation as Jane Street Deal and IPO Talks Emerge

Crusoe closed a $3.9B Series F co-led by Atreides, Mubadala, and Valor at a $30.9B valuation, with Nvidia, Qatar Investment Authority, and Founders Fund participating. A $13B five-year compute contract with Jane Street and IPO conversations with Goldman Sachs and Morgan Stanley underpin the raise.

7d ago release

OpenAI Configures GPT-6 Astra for Legal Research, Scoring 54% vs 38.7% for Base Model

Astra for Law combines GPT-6 Astra with a 230-million-URL legal index and 99.9% of U.S. precedential case law. On the Vals AI Legal Research Bench, it passes 54% of questions versus 38.7% for the base model with web search alone.

7d ago research

Plugin4Shell: Zero-Click RCE Found in Claude Code, Codex, GitHub Copilot, and Gemini CLI

Air Security discloses a SHA-pinning bypass that enables zero-click remote code execution across every major AI coding agent. One design flaw, independently repeated by four labs, leaves millions of developer machines exposed.

7d ago research

DeepMind Institute Maps 11 AGI Economic Interventions Using 51 AI Agent Economists

A new essay from Google DeepMind's research institute evaluates 11 household-facing policy responses to AGI-driven economic disruption using a novel methodology: 51 AI agent raters modeled on real economists. The framework links different interventions to different AGI scenarios rather than backing any single universal policy.

7d ago benchmark

AI Wins the Metaculus Cup for the First Time, Beating Rivals That Spent $15M Combined

On September 5th an AI took first place in the Metaculus Cup forecasting competition, with other AI systems placing second and fifth. Separately, Cassi AI became the first system to outrank human superforecasters on market questions in ForecastBench.

7d ago release

Fujitsu's MONAKA CPU: 2nm, 3.8GHz, Claims 2x AI Inference Throughput with Half the Server Count

Fujitsu begins global sales of the MONAKA CPU in November 2026. Built on a 2nm 3D-stacked process and designed for sovereign AI deployment, the chip is pitched as halving AI inference server requirements versus competing CPUs.

7d ago funding

Hyperscale Data Has Spent $70M to Ready Michigan's 340 MW AI Campus — Operations Start November

Hyperscale Data says it has acquired the equipment needed to bring its Michigan AI data center online in November 2026 under a 20-year master services agreement. The campus holds 340 MW of total potential capacity, with the initial 20 MW now six weeks from first revenue.