TAU2
Measured May 14, 2026Source
Score
0.55
Models13 1 
Claude 4.5 Haiku (Reasoning)
AnthropicAnthropicUnited States
Claude 4.5 Haiku (Reasoning) is a fast, lightweight model from Anthropic optimized for complex reasoning tasks. It combines the speed and cost-efficiency of the Haiku series with enhanced logical and analytical capabilities.
ReasoningFastCheapMultimodal
Input / 1M tokens
$1.25Output / 1M tokens
$5.00Output tokens/s
138.38First-token seconds
6.79sSupported plans
6Benchmark history
Evaluations
Terminalbench Hard
Measured May 14, 2026Source
Score
0.27
Lcr
Measured May 14, 2026Source
Score
0.7
Ifbench
Measured May 14, 2026Source
Score
0.54
Aime 25
Measured May 14, 2026Source
Score
0.84
Scicode
Measured May 14, 2026Source
Score
0.43
Livecodebench
Measured May 14, 2026Source
Score
0.62
Hle
Measured May 14, 2026Source
Score
0.1
Gpqa
Measured May 14, 2026Source
Score
0.67
Mmlu Pro
Measured May 14, 2026Source
Score
0.76
Artificial Analysis Math Index
Measured May 14, 2026Source
Score
83.7
Artificial Analysis Coding Index
Measured May 14, 2026Source
Score
32.6
Artificial Analysis Intelligence Index
Measured May 14, 2026Source
Score
37.1
Plan availability
Products and plans that support this model
Claude Code
Claude Code
Claude Code is Anthropic's agentic coding assistant that works directly in your terminal, IDE (VS Code, JetBrains), web browser, iOS app, and Slack. It reads your codebase, plans and executes multi-file edits, runs tests, and submits pull requests. Uses Claude Opus 4.7, Sonnet 4.6, and Haiku 4.5 models for code understanding and generation.
Discussion
Thinking... Make sure you are connected to GitHub server