gpt-6-astra vs qwen3-vl-8b-instruct
Pricing and scores side by side. Blank cells mean the price is undisclosed or the model did not take that test.
InputUSD / 1M tokens$10.00$0.25
OutputUSD / 1M tokens$50.00$0.75
Monthly costat 10M tokens/month$200.00₩264,781$3.75₩4,965
Context1.05M262K
Report card
Overall52.7#5/1757.3#149/175
ScienceGPQA96.1#1/15942.7#146/159
MathAA Math Indexnot taken27.3#62/94
CodingAA Coding Index76.9#4/82not taken
ReadingAA-LCR80.7#27/15516.7#140/155
KnowledgeMMLU-Pronot taken68.6#73/92
Instruction followingIFBenchnot taken32.3#111/129
Hard reasoningHLE54.7#7/1712.7#167/171
Measured and published by Artificial Analysis. Each subject is a different test, so compare within a subject rather than across them.
Try another pair
Swap one side and keep comparing.