Haiku-5.5 lands #7 on eyebench-v3, scoring +10 over fable-5.1-max at 1/76th the per-task cost
adonis_singh · x · 2026-10-09
The author reports Haiku-5.5 ranks #7 on eyebench-v3, scoring +10 over fable-5.1-max while costing 76x less per task — an impressive price-performance gap.
He adds that Haiku scores the same on the xhigh reasoning setting as on max, which makes the cost difference between fable-5.1 and Haiku on xhigh even more extreme. If accurate, small models at high reasoning effort offer a striking value proposition.
Related event: Haiku-5.5 Ranks 7th on EyeBench-v3 at 1/76 the Cost(2 posts)→
More from Models
- Dev argues for "open-weight models" over "open-source": you can't contribute to them — kipperrii · 2026-10-09
- ChatGPT Invented Court Cases and Lawyers Got Suspended: Inside AI's Legal Hallucination Failures — dadakoglu · 2026-10-09
- User says Qwen3.8 in Hermes "hacked" his PC to prep for CPA exam — natesiggard · 2026-10-09
- Early user: GPT-6 web research feels 10x faster than 5.6 in ChatGPT — flowersslop · 2026-10-09
- FineWeb author: annotating pretraining data with a 27B model is wild but pays off at deployment — antoine_chaffin · 2026-10-09
- HF researcher: fine-tuned small models win on throughput, zero-shot wins on capabilities — antoine_chaffin · 2026-10-09