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 SolHy3
GPT-5.6 Sol

OpenAI

Hy3

Tencent

Pricing
Input $/1M$2.00$0.132
Output $/1M$10.00$0.528
Blended $/1M (3:1)$4.00$0.231
Cache read $/1M$0.20$0.033
Cache write $/1M$2.50
Quality indices
Intelligence Index4725.3
Coding Index77.458.8
Math Index
Agentic Index50.2
Speed & latency
Output speed (tok/s)6584
Time to first token64.46s1.72s
Time to first answer token64.46s25.63s
Benchmarks (Artificial Analysis)
GPQA Diamond94.1%89.7%
Humanity's Last Exam49.5%33.5%
SciCode57.1%48.6%
IFBench72.7%
AA-LCR (Long Context Reasoning)84.0%79.0%
Terminal-Bench Hard65.9%
Terminal-Bench 2.188.0%64.4%
τ²-Bench (Telecom)85.1%
τ³-Bench Banking44.3%22.9%
Design Arena (Elo)
Website1,194
UI components1,179
Data viz1,141
3D1,205
Game dev1,160
Code categories1,192
Specs
Context window1.05M262K
Max output tokens128K128K
Input modalitiesfile, image, texttext
TokenizerGPTOther