mistral-medium-3.5

By price, mistral-medium-3.5 ranks 46th cheapest of 68 on input and 48th on output. Its quality score of 86 ranks 35th of 68, and on value (quality score ÷ output price) it ranks 46th of 68. Its blended rate sits 2% above the catalog average.

In real billing terms: a support chatbot handling 50,000 turns a month (2,000 input / 500 output tokens each) costs about $338 per month on mistral-medium-3.5, while a RAG workload of 20,000 queries (8,000 input / 800 output tokens each) runs about $360. Because output is priced 5.0× input here, long-generation tasks scale cost fastest — cap output length before you optimise anything else.

A 128K context window holds roughly 171 pages of A4 text (at ~750 tokens per page), ranking 53rd of 68 here. That covers typical RAG and multi-turn chat, though book-length inputs still need retrieval and chunking. Output runs at about 90 tok/s (40th fastest of 68), acceptable interactively, though long responses will feel slow.

We track 3 Mistral models, and mistral-medium-3.5's blended rate sits 40% above that provider's average. If your task tolerates a little less headroom, Hunyuan HY3 (Tencent) blends about 97% cheaper at a quality score of 87, and is worth benchmarking first. Going the other way, qwen3.5 scores 91 on quality for about 96% less.

The closest model in Mistral's line-up is mistral-large-3 ($2.00 input / $6.00 output, quality 82). Against it, mistral-medium-3.5 scores 4 points higher on quality, runs 10 tok/s faster (about 13%). When two models sit this close on price, the deciding factor is usually the quality ceiling — run both against your real prompts before picking.