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 Sonnet 5MiniMax M3
Claude Sonnet 5

Anthropic

MiniMax M3

MiniMax

Pricing
Input $/1M$2.00$0.30
Output $/1M$10.00$1.20
Blended $/1M (3:1)$4.00$0.525
Cache read $/1M$0.20$0.06
Cache write $/1M$2.50
Quality indices
Intelligence Index53.444.4
Coding Index71.558.6
Math Index
Agentic Index46.735.4
Speed & latency
Output speed (tok/s)8697
Time to first token93.73s1.29s
Time to first answer token93.73s21.90s
Benchmarks (Artificial Analysis)
GPQA Diamond91.1%92.9%
Humanity's Last Exam39.6%37.1%
SciCode53.6%45.4%
IFBench82.9%
AA-LCR (Long Context Reasoning)70.7%74.0%
Terminal-Bench Hard42.4%
Terminal-Bench 2.180.5%65.2%
τ²-Bench (Telecom)88.9%
τ³-Bench Banking28.2%13.0%
Design Arena (Elo)
Website1,3141,287
Web apps1,3051,263
Full-stack1,2921,250
Mobile apps1,256
Android native1,2551,099
UI components1,3171,280
Data viz1,2811,277
SVG1,2391,227
3D1,3211,280
Game dev1,3451,280
Agentic game dev1,241
Godot game dev1,257
HTML slides1,233
Python→PPTX slides1,252
Code categories1,3121,287
ASCII art1,2301,198
Specs
Context window1M1.05M
Max output tokens128K512K
Input modalitiestext, image, filetext, image, video
TokenizerClaudeOther