Ctrl K
Back to models

Anthropic

Claude Haiku 5.5

  • Released
  • Closed weights
Compare
Intelligence Index

Max setting

43.4#16 of 83measured models
Measured · Artificial Analysis v4.3.2
API price from
$0.10 in / $0.50 out / 1M tokAnthropic · direct API
Context window
1M tok

Performance profile

CodingRank #18 of 94
55.0%
SciCode
Cost efficiencyRank #17 of 67
204.0pts / $
Intelligence ÷ eval cost

Capabilities

Context window

Modalities

Reasoning modes

Avg reasoning tokens / task — Artificial Analysis Intelligence Index v4.3.2: weighted average reasoning tokens per benchmark task, not a token limit. Source (opens in a new tab)

Integration

Reasoning trade-offs

Source (opens in a new tab)
Intelligence Index v4.3.2
Claude Haiku 5.5
Cost per task (USD) · log scale
Claude Haiku 5.5Max
Intelligence Index
43.4
Cost / task
$0.2128
Evaluation time
7.1m

Pricing & access

API route
AnthropicAPI
Tier & context

USD per 1M tokens

Input
$0.10
Output
$0.50
Cached input
$0.01

Available plans · USD

All plans for Claude (opens in a new tab)

Personal

Free
$0/moClaude
Pro
$20/mo
Max 5x
$100/mo
Max 20x
$200/mo
Organizations4 plans
Team Standard
$25/seat/mo
Team Premium
$125/seat/mo
Enterprise
$20/seat/moAnnual · + usage
Education
CustomAccess not verified

Benchmarks

Full leaderboardSource (opens in a new tab)

Humanity’s Last Exam

Frontier reasoning

Claude Haiku 5.5 · Max#15 of 126

  1. 1Claude Opus 5.5Max61.4%
  2. 2Claude Fable 5.1Max59.1%
  3. 3Gemini 4 ArgonHigh57.1%
  5. 12Gemini 3.1 Pro PreviewReasoning47.0%
  6. 13Kimi-K3Max46.9%
  7. 14Step 5 PreviewReasoning46.5%
  8. 15Claude Haiku 5.5Max44.4%
  9. 16Muse Spark OpenExtra high43.9%
  10. 17JT-4.1 Flash 236B A21BReasoning43.6%
  11. 18Grok 4.7Extra high43.1%

Peers use their best measured result on this benchmark.

Alternatives

Each model at its highest-scoring measured setting · Intelligence Index v4.3.2 and its evaluation cost per task.