llama-3.3-70b-instruct vs gpt-6-astra
Pricing and scores side by side. Blank cells mean the price is undisclosed or the model did not take that test.
InputUSD / 1M tokens$0.10$10.00
OutputUSD / 1M tokens$0.32$50.00
Monthly costat 10M tokens/month$1.55₩2,052$200.00₩264,781
Context131K1.05M
Report card
Overall7.7#145/17552.7#5/175
ScienceGPQA49.8#136/15996.1#1/159
MathAA Math Index7.7#84/94not taken
CodingAA Coding Index11.9#75/8276.9#4/82
ReadingAA-LCR15.7#141/15580.7#27/155
KnowledgeMMLU-Pro71.3#67/92not taken
Instruction followingIFBench47.1#67/129not taken
Hard reasoningHLE3.6#157/17154.7#7/171
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.