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 4.5GPT-5.5
Claude Haiku 4.5

Anthropic

GPT-5.5

OpenAI

Pricing
Input $/1M$1.00$5.00
Output $/1M$5.00$30.00
Blended $/1M (3:1)$2.00$11.25
Cache read $/1M$0.10$0.50
Cache write $/1M$1.25
Quality indices
Intelligence Index29.654.8
Coding Index43.974.9
Math Index
Agentic Index16.444.9
Speed & latency
Output speed (tok/s)92
Time to first token40.65s
Time to first answer token40.65s
Benchmarks (Artificial Analysis)
GPQA Diamond93.5%
Humanity's Last Exam44.3%
SciCode56.1%
IFBench75.9%
AA-LCR (Long Context Reasoning)74.3%
Terminal-Bench Hard60.6%
Terminal-Bench 2.184.3%
τ²-Bench (Telecom)93.9%
τ³-Bench Banking31.3%
Design Arena (Elo)
Website1,1491,280
Web apps1,171
Full-stack1,142
Mobile apps1,218
Android native1,241
UI components1,1421,294
Data viz1,1601,293
SVG1,0721,273
3D1,1311,250
Game dev1,1551,343
Agentic game dev1,191
Godot game dev1,204
HTML slides1,083
PPTX slides1,157
Python→PPTX slides1,152
Agentic slides1,150
Agentic HTML slides1,084
Agentic slides (HTML)1,077
Agentic slides (Python PPTX)1,155
Code categories1,1481,284
ASCII art1,1821,288
Specs
Context window200K1.05M
Max output tokens64K128K
Input modalitiestext, image, filefile, image, text
TokenizerClaudeGPT