Artificial Analysis found MiniMax M3.1 Flash generated approximately 189 million output tokens during its Intelligence Index evaluation, compared to roughly 324 million for GLM-5.3 Flash and 334 million for Gemini 3.8 Flash.

Overview Debates Related claims Topics Sources

From the articleMiniMax's new M3.1 Flash beats DeepSeek and Opus 5 on a key business benchmark at lower cost — but with a high hallucination rate

441e85161Article

MiniMax's new M3.1 Flash beats DeepSeek and Opus 5 on a key business benchmark at lower cost — but with a high hallucination rate