🧰 GPT-5.6 Luna

The items that name it 9

• How we detect frontier model misalignment
  articleas tools2026-10-29

  Monitor model compared in monitoring evaluations

• ReSI: Recursive Safety Improvement toward Resistant and Resilient AI
  paperas affected products2026-10-08It is written here as GPT-5.6-Luna (frontier model; leading frontier result of 56.69% mean X-Teaming attack success rate).

  frontier model; leading frontier result of 56.69% mean X-Teaming attack success rate

• GPT-6 Sol and GPT-6 Luna: October 2026 update
  articleas affected products2026-10-07It is written here as GPT-5.6 Luna (August).

  August

• Top AI agent security resources — October 2026
  articleas affected products2026-10-04It is written here as GPT-5.6-Luna (OpenAI).

  OpenAI

• Addendum to GPT-6 Astra System Card: GPT-6.1 Sol
  articleas affected products2026-09-29
• Building a Faster, Cheaper Vulnerability Scanner With Jev | depthfirst
  articleas tools2026-09-25

  LLM used as the comparison baseline: input costs five times more than Jev, calls take around 10 seconds, context window is 1 million tokens, and scans take at least 3 minutes with most other models.

• AI for Web3 Security: Zero to Hero
  courseas tools2026-09-21It is written here as gpt-5.6-luna.

  model shipped as default in lucid/llm.py, used to produce the course's baseline runs, routed through Requesty

• We burned 11.7bn tokens to find the best cyber AI model
  articleas tools2026-08-21

  AI model benchmarked for vulnerability rediscovery; surpassed by GLM-5.3 final release

• GPT-5.6 — August Updates
  writeupas affected products2026-08-06

In every workspace you can read 36