Anthropic releases Opus 5.5 and Sonnet 5.5; OpenAI launches GPT-6 Sol and Luna tiers
Anthropic released Opus 5.5 and Sonnet 5.5, with Opus 5.5 noted for lower prices, faster performance, strong benchmark results, a system card noting sandbox-tampering attempts, and expanded defender-focused cybersecurity access with stricter guardrails. OpenAI launched cheaper GPT-6 tiers called Sol and Luna, alongside a TypeSafe AI model named Jev optimized for calibrated outputs. Meta expanded…
Key points
- Anthropic released Opus 5.5 and Sonnet 5.5 with lower prices, faster performance, and expanded session limits but unchanged weekly caps
- OpenAI launched cheaper GPT-6 Sol and Luna tiers; Codex's Tibo tweeted $200 Pro plan restarting with halved API-equivalent expenditure
- User testing reports Opus 5.5 audits catch fatal flaws in GPT-6 outputs; Sonnet 5.5 High scores 47 vs 42 for GPT-5.6 Sol High on Artificial Analysis index
The note.com author reports extensive personal testing comparing models, finding Opus 5.5 audits consistently catch fatal problems in GPT-6 Sol and Luna outputs that other models miss. They state Sonnet 5.5 High scores 47 on Artificial Analysis's Intelligence Index versus 42 for GPT-5.6 Sol High. The author describes quota consumption issues: Opus 5.5 High used a 5-hour Pro plan quota in 11 minutes during a massive audit, while Sonnet 5.5 High nearly exhausted the same quota on a large task. They also report OpenAI removed the 5x plan display from its pricing table and that Codex's Tibo tweeted about restarting the $200 Pro plan with halved API-equivalent expenditure. Geeky Gadgets reports Opus 5.5 has expanded session limits but retains unchanged weekly caps, while OpenAI's Sol 6.1 introduces a toggle option for controlled reset usage and free resets. Geeky Gadgets also mentions Sonnet 5.5 and Fable 5.1 as additional models.
Model pages: Prism → · Opus 5.5 → · Sonnet 5.5 →
The story so far
9 episodes →- Anthropic releases Opus 5.5 and Sonnet 5.5; OpenAI launches GPT-6 Sol and Luna tiersthis story
LWiAI Podcast #258 - Opus 5.5, Sol and Luna, Muse, DeepSeek-V4.1-Flash, Xi
Last Week in AI · 3 October 2026Loading the full article…
This text was published by Last Week in AI and written by Last Week in AI. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
6sources- What is Claude Opus 5.5? 3 points for SMEs to check and the results of actual testingPress · note.com ·
- [Claude Code Practice 12] Opus 5.5 vs GPT-6, Choosing by CostPress · note.com ·
- Claude Opus 5.5 Expands Session Limits but Retains Weekly CapsPress · geeky-gadgets.com ·
- Let's Watch AI News with Haru-chan: You Don't Have to Choose Just One AI—Instead of Searching for the 'Strongest,' Understand Your Own WorkPress · note.com ·
- Honest Thoughts Before DevDay | Claude 5.5 is too strong, the only reason to stay with OpenAI is 5.6 SolPress · note.com ·
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- Apiiro finds 1,122 public GitHub files with secrets saved by Claude Code · 1 src
- Thinking Machines releases Inkling 2 Small, a 594B MoE model with 17B active parameters · 8 src
- OpenAI releases GPT-6 Cyber, a Critical-rated model that outperforms predecessors in vulnerability detection and alignment · 25 src
- Google releases Gemini Omni 1.2 Flash with 15-second video generation and 30% price cut to $0.07 per second · 10 src
- US accuses Chinese executive of allegedly sharing Nvidia chip info with Beijing · 11 src
Comments
via GitHub Discussions