Mistral Medium 3.5 is a dense 128B instruction-following model from Mistral AI. It supports text and image inputs with text output, and is designed for agentic workflows, coding, and complex...
Tested on 21 benchmarks · BenchGecko score 25.0. Top scores: Chatbot Arena Elo — Overall (1426.9%), Chatbot Arena Elo — Coding (1262.9%), Artificial Analysis · tau2-Bench Telecom (94.2%).
Command R (08-2024) scores 26.2 (105% as good) at $0.15/1M input · 90% cheaper
Unusual and adversarial machine learning challenges. Tests robustness of reasoning about edge cases in ML systems.
Chatbot Arena overall Elo rating. Crowdsourced human preference ranking from blind head-to-head comparisons across all topics.
Chatbot Arena coding Elo. Human preference ranking specifically for coding tasks and technical questions.
- Typemultimodal
- Context262K tokens (~131 books)
- ReleasedApr 2026
- LicenseOpen Source
- StatusActive
- Cost / Message~$0.010
Frequently Asked Questions
Related Models
CodeQwen1.5-7B · Alibaba QwenLlama 2 7b Chat Hf · MetaClaude 3.5 Haiku · AnthropicMixtral 8x22B Instruct · Mistral AIQwen2 1.5B Instruct · AlibabaKey facts · as of 2026-10-05
- Mistral Medium 3.5 by Mistral AI. BenchGecko score 25.0, rank 265 of 312 scored models (normalized average of public benchmark scores).
- List price $1.50 input · $7.50 output per 1M tokens (as of 2026-10-05).
- Gecko Tests: Who Are You E (Sometimes says it is OpenAI) · Tokenizer Tax C (53% more tokens outside English).
How to cite · data as of 2026-10-05
Mistral Medium 3.5 · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-05. https://benchgecko.ai/model/mistral-medium-3-5
Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP