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.

GPT-5.6 SolQwen2.5 72B Instruct
GPT-5.6 Sol

OpenAI

Qwen2.5 72B Instruct

Alibaba (Qwen)

Pricing
Input $/1M$5.00$0.36
Output $/1M$30.00$0.40
Blended $/1M (3:1)$11.25$0.37
Cache read $/1M$0.50
Cache write $/1M$6.25
Quality indices
Intelligence Index58.99.6
Coding Index77.4
Math Index14
Agentic Index54
Speed & latency
Output speed (tok/s)670
Time to first token120.38s0s
Time to first answer token120.38s0s
Benchmarks (Artificial Analysis)
MMLU-Pro72.0%
GPQA Diamond94.1%49.1%
Humanity's Last Exam47.2%4.2%
LiveCodeBench27.6%
SciCode56.1%26.7%
MATH-50085.8%
AIME16.0%
AIME 202514.0%
IFBench72.7%36.9%
AA-LCR (Long Context Reasoning)73.7%20.3%
Terminal-Bench Hard65.9%4.5%
Terminal-Bench 2.188.0%
τ²-Bench (Telecom)85.1%34.5%
τ³-Bench Banking33.0%
Specs
Context window1.05M33K
Max output tokens128K16K
Input modalitiesfile, image, texttext
TokenizerGPTQwen