2026 in LLMs (So Far) — Simon Willison

Epoch Capabilities Index

Epoch AI's index combining many benchmarks into a single capability scale for comparing AI models over time, with a release-date graph view.

Arena AI

A public leaderboard where visitors chat with, compare and vote on AI models, producing community rankings for language, image and code models based on real-world use.

Visual Explainer (Agent Skill)

An open-source agent skill on GitHub by nicobailon that generates rich HTML pages or slide decks for diagrams, diff reviews, plan audits, data tables, and project recaps.

Upstage Releases Solar Mini 4 (Artificial Analysis)

Artificial Analysis on Solar Mini 4, a proprietary reasoning model from Korean lab Upstage. It scores 24 on the Artificial Analysis Intelligence Index, with Upstage reporting 35B total and 3B active parameters, and is priced at $0.10/$0.40 per million input/output tokens — though it costs about 5x as much per task as GPT-6 Luna (max).

OWASP Gen AI Security Project

Community-led OWASP project publishing free, open-source guidance on security and safety risks in generative AI applications, including the Top 10 for LLM applications.

SkillGym: Internalizing Human Skills into LLMs for Real-World Problem Solving

Authors

Zhilong Ge, Yuting Shao, Yutao Yang, Yuxuan Cai, Jie Zhou, et al.

Venue

Preprint, not yet peer-reviewed

Published

September 2026

ID

arXiv:2609.27717

arXiv paper (Sept. 2026) introducing SkillGym, a framework that turns human-written agent skills into executable, verifiable training environments so language model agents learn them as built-in capabilities.