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 Opus 4.6MiMo-V2.5
Claude Opus 4.6

Anthropic

MiMo-V2.5

Xiaomi

Pricing
Input $/1M$5.00$0.14
Output $/1M$25.00$0.28
Blended $/1M (3:1)$10.00$0.175
Cache read $/1M$0.50$0.0028
Cache write $/1M$6.25
Quality indices
Intelligence Index37.837.2
Coding Index56.8
Math Index
Agentic Index23.7
Speed & latency
Output speed (tok/s)5763
Time to first token1.85s3.40s
Time to first answer token1.85s34.93s
Benchmarks (Artificial Analysis)
GPQA Diamond84.0%84.9%
Humanity's Last Exam18.6%25.2%
SciCode45.7%43.1%
IFBench44.6%67.1%
AA-LCR (Long Context Reasoning)58.3%62.7%
Terminal-Bench Hard48.5%41.7%
Terminal-Bench 2.163.7%
τ²-Bench (Telecom)84.8%90.6%
τ³-Bench Banking6.6%
Design Arena (Elo)
Website1,3231,288
Web apps1,261
Full-stack1,279
Mobile apps1,280
Android native1,195
UI components1,3361,294
Data viz1,3221,279
SVG1,2741,218
3D1,3351,273
Game dev1,3371,291
Code categories1,3271,286
ASCII art1,2981,175
Specs
Context window1M1.05M
Max output tokens128K131K
Input modalitiestext, image, filetext, audio, image, video
TokenizerClaudeOther