llama-3.1-70b-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$0.40$10.00
OutputUSD / 1M tokens$0.40$50.00
Monthly costat 10M tokens/month$4.00₩5,296$200.00₩264,781
Context131K1.05M
Report card
Overall6.6#164/17552.7#5/175
ScienceGPQA40.9#150/15996.1#1/159
MathAA Math Index4.0#88/94not taken
CodingAA Coding Indexnot taken76.9#4/82
ReadingAA-LCRnot taken80.7#27/155
KnowledgeMMLU-Pro67.6#78/92not taken
Instruction followingIFBench34.4#105/129not taken
Hard reasoningHLE4.5#137/17154.7#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.