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 4.5Step 3.7 Flash
Claude Haiku 4.5

Anthropic

Step 3.7 Flash

Stepfun

Pricing
Input $/1M$1.00$0.20
Output $/1M$5.00$1.15
Blended $/1M (3:1)$2.00$0.438
Cache read $/1M$0.10$0.04
Cache write $/1M$1.25
Quality indices
Intelligence Index16.919.5
Coding Index43.939.6
Math Index
Agentic Index8
Speed & latency
Output speed (tok/s)179
Time to first token1.39s
Time to first answer token12.59s
Benchmarks (Artificial Analysis)
GPQA Diamond80.9%
Humanity's Last Exam21.4%
SciCode43.9%
IFBench67.3%
AA-LCR (Long Context Reasoning)73.7%
Terminal-Bench Hard35.6%
Terminal-Bench 2.139.3%
τ²-Bench (Telecom)98.5%
τ³-Bench Banking12.0%
Design Arena (Elo)
Website1,1351,207
UI components1,1151,198
Data viz1,1391,185
SVG1,0491,099
3D1,1011,157
Game dev1,1201,181
Code categories1,1301,196
ASCII art1,1591,167
Specs
Context window200K262K
Max output tokens64K230K
Input modalitiestext, image, filetext, image, video
TokenizerClaudeOther