HomeTagsBenchmarks

Benchmarks

Stories from Benchmarks come first, newest first, then stories from connected topics (AI Models, LLM, AI, coding agents, preprint, open weights), each labelled with where it comes from. This view is not indexed by search engines.

MLPerf Training v6.1 Adds First Rubin GPUs and TPUv7

MLCommons has published MLPerf Training v6.1 results from 153 systems, including the first NVIDIA Rubin GPUs and Google TPUv7.

Just in

Latest

Anthropic Launches Claude Fable 5.2 and Mythos 5.2

Anthropic launched Claude Fable 5.2 for all users and Mythos 5.2 via verification. Benchmarks are Anthropic’s; Opus 5.5 stays the daily driver.

Meta Releases Muse Ember, Its Smallest Agent Model

Meta released Muse Ember, its smallest Muse model for agent work, at $0.10/$0.50 per million tokens. Benchmarks are Meta’s; Haiku leads FrontierCode.

FMF Jailbreak Rating Framework Opens for Comment to 21 Dec

FMF Jailbreak Rating Framework v1.0 rates cyber jailbreaks on four dimensions. Voluntary; comment open to 21 Dec 2026. Not law or certification.

Ricursive Releases Ripple-65B Preview Open Weights

Ricursive Intelligence released Ripple-65B Preview open weights. It claims 80.52 on TerminalBench 2.1 among US peers and a $32M R&D cost.

Meituan’s LongCat-2.5 AI Model Was Trained Only on Chinese Chips

Meituan's LongCat-2.5 is a 1.6tn-parameter multimodal model trained on Chinese chips. Its own benchmarks show wins and clear losses against Claude.

AI Reasoning Models Diagnose Better but Add 34% More False Positives

AI reasoning models made more accurate diagnoses but over 34% more false positives in a retrospective study; doctors checking them cut those by 44%.

Microsoft Open-Sources Mage-VL-Cowork, a 4B Agent That Clicks or Codes

Microsoft's open Mage-VL-Cowork is a 4B computer-use agent that switches between clicking and code. Microsoft reports 66% on OSWorld-Verified.

Factory Says Droid 1.0 Beats Codex and Claude Code in Its Own Tests

Factory's Droid 1.0 coding-agent harness makes scripted tool calls the default. Its own Terminal-Bench 4 runs put it ahead of Codex and Claude Code.

MLCommons Ranks 27 AI Systems on Safety; Open Models Span 4th to 98th

MLCommons' AILuminate Safety 1.2.1 puts 22 open-weight models between the 4th and 98th percentile. Jailbreak 1.1 scores follow after developers are told.

Mistral Opens Large 4 Weights: a 1-Trillion-Parameter Model

Mistral has released the weights of Large 4, a 1-trillion-parameter multimodal model, and warns its cyber skills carry a potential for real attack uplift.

Latest on TSN

Explore the TSN Atlas: Every topic we cover, mapped. Click a bubble to dive in.