Mistral Large 4 coding benchmark

How Mistral Large 4 performed

Mistral Large 4 by Mistral AI took the PomBench AI coding benchmark: every model gets the same prompt and writes self-contained browser code for a twelve-second scene: a Pomeranian runs across a backyard, jumps a puddle, enters its doghouse, retrieves a bone and turns toward the viewer. The video above is an independent offline browser recording of the code Mistral Large 4 returned, captured at 960 × 540 and 30 fps, with no organizer edits to the code.

In this Mistral Large 4 coding test the model produced 175,024 output tokens, including provider-reported reasoning (rank 1 of 46 by output tokens), and the known API cost was $0.7578 (rank 43 of 45, cheapest first). Is Mistral Large 4 good at coding? One task cannot answer that on its own, so this LLM coding benchmark lets you watch the result instead of reading a score: check the gait, the puddle jump and reflection, the doorway occlusion and whether the bone stays in the dog's mouth. This run was served through OpenRouter with one fixed upstream provider and fallbacks disabled.

Then compare Mistral Large 4 vs Mistral Medium 3.5 and Mistral Large 3 below, or open the AI model comparison to play up to four models on one synchronized timeline.