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.

DeepSeek V3GPT-5.6 Sol
DeepSeek V3

DeepSeek

GPT-5.6 Sol

OpenAI

Pricing
Input $/1M$0.20$5.00
Output $/1M$0.80$30.00
Blended $/1M (3:1)$0.35$11.25
Cache read $/1M$0.50
Cache write $/1M$6.25
Quality indices
Intelligence Index14.258.9
Coding Index2377.4
Math Index26
Agentic Index54
Speed & latency
Output speed (tok/s)070
Time to first token0s112.90s
Time to first answer token0s112.90s
Benchmarks (Artificial Analysis)
MMLU-Pro75.2%
GPQA Diamond55.7%94.1%
Humanity's Last Exam3.6%47.2%
LiveCodeBench35.9%
SciCode35.4%56.1%
MATH-50088.7%
AIME25.3%
AIME 202526.0%
IFBench34.8%72.7%
AA-LCR (Long Context Reasoning)29.0%73.7%
Terminal-Bench Hard6.8%65.9%
Terminal-Bench 2.116.9%88.0%
τ²-Bench (Telecom)22.8%85.1%
τ³-Bench Banking4.7%33.0%
Design Arena (Elo)
Website1,147
UI components1,136
Data viz1,135
SVG1,023
3D1,146
Game dev1,112
Code categories1,142
Specs
Context window164K1.05M
Max output tokens16K128K
Input modalitiestextfile, image, text
TokenizerDeepSeekGPT