claude-haiku-5-5

Common Name: Claude Haiku 5.5

Anthropic
Released on Oct 7 12:00 AMKnowledge Cutoff Jun 1 12:00 AMSupportedTool InvocationSupportedReasoning
CompareTry in Chat

Anthropic's fastest and most affordable small model for high-volume, latency-sensitive work such as classification, routing, extraction, and subagents, with a 1M token context window.

Specifications

Context
1000K
Maximum Output
128K
Inputtext, image, pdf
Outputtext

Performance (7-day Average)

Collecting…
Collecting…
Collecting…

Availability Trend (24h)

Pricing

< 100K
Input$0.11/MTokens
Output$0.55/MTokens
Batch Input$0.055/MTokens
Batch Output$0.275/MTokens
Cache Creation$0.1375/MTokens
text.input.cache creation 1h$0.22/MTokens
Cached Input$0.011/MTokens
100K+
Input$0.55/MTokens
Output$2.75/MTokens
Batch Input$0.275/MTokens
Batch Output$1.375/MTokens
Cache Creation$0.6875/MTokens
text.input.cache creation 1h$1.10/MTokens
Cached Input$0.055/MTokens

Performance Metrics (24h)

Similar Models

Claude Opus 4.7

$5.50/$27.50/M
ctx1.0Mmax128Kavail—tps—
InOutCap
Knowledge#9

Anthropic's latest Opus model with stronger advanced software engineering and higher-resolution vision.

Claude Fable 5

$11.00/$55.00/M
ctx1.0Mmax128Kavail—tps—
InOutCap
Reasoning#2Scicode#2

Anthropic's most capable widely released Mythos-class model for demanding reasoning, long-horizon agentic work, software engineering, knowledge work, vision, and scientific research.

Claude Fable 5.1

$11.00/$55.00/M
ctx1.0Mmax128Kavail—tps—
InOutCap
Reasoning#1Knowledge#1

Claude Fable 5.1 extends Claude Fable 5 with stronger long-running agentic coding, knowledge work, and research, at the same input/output prices with cache reads a quarter of the cost.

Claude Opus 5.5

$4.40/$22.00/M
ctx1.0Mmax128Kavail—tps—
InOutCap

Anthropic's Claude Opus 5.5 for long-running agentic coding and knowledge work, cheaper per token than Opus 5 with adaptive thinking always on.