HomeTagsBenchmarks

Benchmarks - Page 16

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.

Latest

GitLab Orbit Goes GA, Claiming Up to 45x Fewer Agent Retries

GitLab Orbit, a context graph for AI coding agents, is now GA. GitLab says its tests showed up to 45x fewer retries and 4.5x fewer tokens.

Kakao Open-Sources Two Kanana-2 Embedding Models for Korean Search

Kakao has released 3bn and 1.3bn-parameter Kanana-2 embedding models on Hugging Face, topping its Korean retrieval tests while trailing Qwen3 in English.

Particle-1 Coding Model Uses About 45% Fewer Tokens, Particle Says

Suhail Doshi's Particle has launched Particle-1, a DeepSeek V4.1 Flash coding model it says uses about 45% fewer tokens with similar accuracy.

UCLA’s MAESTRO AI Ranked 3,592 Compounds to Pair With a Cancer Drug

UCLA's MAESTRO uses large language models to rank drug pairs. It flagged berberine with enzalutamide, which worked together in lab prostate cancer models.

SenseTime Releases Code2Video, an Open Model That Makes Videos in Code

SenseTime's open SenseNova 6.8 Code2Video, a 27B model built on Qwen3.8-27B, writes and renders code to make short videos. It has no benchmarks yet.

Microsoft Opens Its Copilot Agent Harness to Developers via Foundry

Microsoft is offering the harness behind GitHub Copilot and Copilot Studio in Foundry, in preview, free until billing starts on 1 December 2026.

Replit Cloud Agents Let Non-Engineers Send Tested Pull Requests

Replit says its Cloud Agents let staff outside engineering describe a code change and get a tested pull request with video, while engineers keep approval.

Screenshot to UI Code: The Tools That Actually Do It

Only two of five viral 'screenshots to UI' GitHub repos do the job. We check each, add v0, Bolt.new, Lovable and Napkins, and flag a GPT Pilot worm.

Cursor’s Agent SDK Turns AI Agents Into Code Projects You Can Test

Cursor's Agent SDK defines AI agents as Markdown and TypeScript projects with built-in evals. How it differs from Grok Bot and Cursor Automations.

Microsoft’s MAI Image 2.7 Preview Ranks in Top Three for Image Editing

Microsoft AI's MAI Image 2.7 editing model, now in preview, ranks third on Arena's editing leaderboard and second on Artificial Analysis.

Latest on TSN

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