Claude Haiku 5.5 vs Claude Opus 5.5

Pyyan / Compare / Claude Haiku 5.5 vs Claude Opus 5.5

2 of 5

Language Models · verified 10 Oct 2026

Claude Haiku 5.5Anthropiccurrent

Claude Opus 5.5Anthropiccurrent

Add another3 slots left

Specification	Claude Haiku 5.5	Claude Opus 5.5
Summary	A tenth of Haiku 4.5's price, and it finishes terminal tasks Haiku 4.5 could not start.	Fable 5.1's level of work on most tasks, at 40% less to run than Opus 5.
Context	Not published	1M
Input ($/Mtok)	$0.1	$4
Output ($/Mtok)	$0.5	$20
Licence	Proprietary	Proprietary
Max output	Not published	128K
Parameters	Not published	Not published
Released	7 Oct 2026	22 Sep 2026
Model ID	claude-haiku-5-5	claude-opus-5-5
Category	Language Models	Language Models
Official	Anthropic ↗	Anthropic ↗

Highlighted rows are where these differ.

39.2% on Terminal-Bench 4.0, where Haiku 4.5 scored 0.0%
Above 100K tokens the price rises to $0.50 and $2.50, for the whole request
First Haiku with an adjustable effort setting, and a new tokenizer that spends slightly more tokens

Best for high volume work at the bottom of the stack.

66.4% on Terminal-Bench 4.0 against 52.3% for Opus 5, on Anthropic's own harness
Cache reads cut 60% to $0.20, and output more than 30% faster than Opus 5
Attempts to work around its boundaries 85% less often than Opus 5, Anthropic reports

Best for agentic coding at flagship quality.