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 4.6DeepSeek V3.1
Claude Opus 4.6

Anthropic

DeepSeek V3.1

DeepSeek

Pricing
Input $/1M$5.00$0.25
Output $/1M$25.00$0.95
Blended $/1M (3:1)$10.00$0.425
Cache read $/1M$0.50$0.13
Cache write $/1M$6.25
Quality indices
Intelligence Index37.821
Coding Index
Math Index49.7
Agentic Index
Speed & latency
Output speed (tok/s)540
Time to first token1.89s0s
Time to first answer token1.89s0s
Benchmarks (Artificial Analysis)
MMLU-Pro83.3%
GPQA Diamond84.0%73.5%
Humanity's Last Exam18.6%6.3%
LiveCodeBench57.7%
SciCode45.7%36.7%
AIME 202549.7%
IFBench44.6%37.8%
AA-LCR (Long Context Reasoning)58.3%45.0%
Terminal-Bench Hard48.5%24.2%
τ²-Bench (Telecom)84.8%34.8%
Design Arena (Elo)
Website1,3231,150
Web apps1,261
Full-stack1,279
Mobile apps1,280
Android native1,195
UI components1,3361,126
Data viz1,3221,136
SVG1,2741,014
3D1,3351,136
Game dev1,3371,143
Code categories1,3271,145
ASCII art1,298
Specs
Context window1M164K
Max output tokens128K33K
Input modalitiestext, image, filetext
TokenizerClaudeDeepSeek