Anthropic releases Claude Haiku 5.5, cutting prices by 75%

Anthropic released Claude Haiku 5.5 on October 7, 2026, pricing input tokens at $0.10 per million and output tokens at $0.50 per million for prompts up to 100,000 tokens, with rates of $0.50 and $2.50 above that threshold. Anthropic said the model costs around 75% less to run on average than its predecessor, Haiku 4.5.

The published benchmark table claims step-changes over the prior Haiku generation. On GDPval-AA v2.1, a knowledge-work evaluation, Haiku 5.5 scores 1620 Elo against 735 for Haiku 4.5. On the offline subset of OSWorld 2.1, a computer-use benchmark, it posts 72.4% versus 15.7%. Humanity's Last Exam, a test of expert-level academic reasoning, comes in at 45.9% without tools and 57.4% with tools, up from 10.2% and 18.7% on the predecessor.

Anthropic routed much of its launch write-up through named customer quotes. Asana reported "over a 30% reduction in latency" and "up to 2.5x faster inference." AlphaSense saw a "statistically significant improvement" on document Q&A, moving from 0.76 to 0.84. Box said Haiku 5.5 scored "11 points higher than Haiku 4.5 at about half the latency." HubSpot cited 92.8% averaged over three runs on CRM evaluations.

The model is available on AWS, Google Cloud, and Microsoft Azure under the ID claude-haiku-5-5. Anthropic says its cybersecurity safeguards permit defensive tasks while blocking penetration testing. Three analysts in our Who's Who tracker shared the launch post the same day it dropped.