AiCostCompare

Compare models

Pick up to 4 models for a side-by-side spec sheet. Hover any dotted metric name for what the test measures. Best value in each row is highlighted.

Claude Haiku 4.5Llama 4 Scout
Claude Haiku 4.5

Anthropic

Llama 4 Scout

Meta

Pricing
Input $/1M$1.00$0.10
Output $/1M$5.00$0.30
Blended $/1M (3:1)$2.00$0.15
Cache read $/1M$0.10
Cache write $/1M$1.25
Quality indices
Intelligence Index29.610
Coding Index43.98.2
Math Index14
Agentic Index16.41.1
Speed & latency
Output speed (tok/s)76
Time to first token0.56s
Time to first answer token0.56s
Benchmarks (Artificial Analysis)
MMLU-Pro75.2%
GPQA Diamond58.7%
Humanity's Last Exam4.3%
LiveCodeBench29.9%
SciCode17.0%
MATH-50084.4%
AIME28.3%
AIME 202514.0%
IFBench39.5%
AA-LCR (Long Context Reasoning)25.8%
Terminal-Bench Hard1.5%
Terminal-Bench 2.13.7%
τ²-Bench (Telecom)15.5%
τ³-Bench Banking3.3%
Design Arena (Elo)
Website1,149778
UI components1,142811
Data viz1,160933
SVG1,072
3D1,131
Game dev1,155831
Code categories1,148823
ASCII art1,182
Specs
Context window200K1.31M
Max output tokens64K16K
Input modalitiestext, image, filetext, image
TokenizerClaudeLlama4