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 4.6Step 3.7 Flash
Claude Opus 4.6

Anthropic

Step 3.7 Flash

Stepfun

Pricing
Input $/1M$5.00$0.20
Output $/1M$25.00$1.15
Blended $/1M (3:1)$10.00$0.438
Cache read $/1M$0.50$0.04
Cache write $/1M$6.25
Quality indices
Intelligence Index26.419.5
Coding Index39.6
Math Index
Agentic Index
Speed & latency
Output speed (tok/s)0179
Time to first token0s1.39s
Time to first answer token0s12.59s
Benchmarks (Artificial Analysis)
GPQA Diamond84.0%80.9%
Humanity's Last Exam19.1%21.4%
SciCode43.9%
IFBench44.6%67.3%
AA-LCR (Long Context Reasoning)67.0%73.7%
Terminal-Bench Hard48.5%35.6%
Terminal-Bench 2.139.3%
τ²-Bench (Telecom)84.8%98.5%
τ³-Bench Banking12.0%
Design Arena (Elo)
Website1,3031,207
Web apps1,219
Full-stack1,229
Mobile apps1,238
Android native1,216
UI components1,3001,198
Data viz1,2961,185
SVG1,2511,099
3D1,3031,157
Game dev1,3011,181
Agentic game dev1,215
Code categories1,3041,196
ASCII art1,2731,167
Specs
Context window1M262K
Max output tokens128K230K
Input modalitiestext, image, filetext, image, video
TokenizerClaudeOther