Claude Haiku 5.5 ships with huge jumps: OSWorld 15.7%→72.4%, beats GPT-6 Luna across the board
mark_k · x · 2026-10-08
Anthropic's smallest model Claude Haiku 5.5 just dropped with a massive benchmark leap, beating GPT-6 Luna on every chart where both are reported: OSWorld 2.1 computer use 15.7%→72.4% (vs Luna's 48.9%), HLE without tools 10.2%→45.9% (57.4% with tools), Terminal-Bench 4.0 0%→39.2% (vs Luna's 16.4%), FrontierCode 46.4%, Chartography 6.4%→46.4%, and knowledge-work scores of 1620 GDPval-AA / 1578 AA-Briefcase. Sonnet 5.5 still leads overall, but the generational jump — especially in computer use and reasoning — is enormous.
Related event: Anthropic Launches Claude Haiku 5.5 with ~75% Lower Running Costs(59 posts)→
More from Models
- 'Compression of the RH proof has begun': AI circles buzz over unverified Riemann claim — teortaxesTex · 2026-11-22
- GLM 5.3 and DeepSeek V4.1 Flash resold at 60% off: the economics of open-source giveaways — lxfater · 2026-11-22
- GPT-6.2 Astra and Sol answer more directly but score lower on transparency than Claude — The Decoder · 2026-11-22
- Baseline model test: no persona, no forced formatting, no safety bypasses observed — Sauers_ · 2026-11-22
- 2-day Gemini stress test: $4.99 Google AI Pro nearly maxed, orchestrator UI revealed — lipeng0820 · 2026-11-22
- Timed eval task asks users to deliberately make Claude fail at real work in 12 minutes — Icy-Measurement10 · 2026-11-22