llama-3.1-8b-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$0.05$10.00
OutputUSD / 1M tokens$0.08$50.00
Monthly costat 10M tokens/month$0.58₩761$200.00₩264,781
Context131K1.05M
Report card
Overall6.9#155/17552.7#5/175
ScienceGPQA25.9#157/15996.1#1/159
MathAA Math Index4.3#87/94not taken
CodingAA Coding Index5.4#81/8276.9#4/82
ReadingAA-LCR18.0#138/15580.7#27/155
KnowledgeMMLU-Pro47.6#89/92not taken
Instruction followingIFBench28.6#121/129not taken
Hard reasoningHLE5.3#127/17154.7#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.