Up to four models
DeepSeek-V3 vs Mistral Small 4
On the index (GPT-4 = 100), DeepSeek-V3 scores 119 and Mistral Small 4 119. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
119103–135
WeightsOpen
ReleasedDec 2024
Price in / out$0.26 / $1.03
Context164K
Core tests taken2 of 9
119100–138
WeightsOpen
ReleasedMar 2026
Price in / out$0.15 / $0.60
Context262K
Core tests taken0 of 9
DeepSeek-V3 and Mistral Small 4 are too close to call (0 points apart, within the uncertainty), at 1.7× the price.
Test by testthird-party results
These models share no third-party tests yet.
Task length · METR
28 sec
Tests at least two of these models have taken. Bold rows are the core tests.