Claude Haiku 5.5 matches GPT-6 Luna pricing but new tokenizer hides a 25% increase
RexDouglass · x · 2026-10-08
- Anthropic released Claude Haiku 5.5: $0.10/$0.50 per million tokens up to 100K tokens, exactly matching GPT-6 Luna; beyond 100K it jumps 5x to $0.50/$2.50, while Luna only rises to $0.20/$0.75 at 272K.
- Hidden cost: the new tokenizer consumes 1.25x more tokens on the same long prompt, an effective 25% price increase.
- Verdict: under 100K tokens Haiku matches Luna on price with higher benchmarks; for long contexts Luna is the better deal.
- Willison also benchmarked reasoning levels with pelican SVG tests: low effort took 7s at 0.09 cents, max effort 5m9s at 3.4 cents; reasoning can't be disabled and defaults to medium.
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