Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
Pioneer logo

Model details

Mistral Medium 3

Mistral Medium 3 is an enterprise-class language model unveiled by Mistral AI on May 7, 2025 through the official blog post "Medium is the new large." The release introduces a new tier positioned between Mistral's openly published models (such as Mistral Small and Mistral Large) and its narrower enterprise offerings (such as Mistral OCR, Mistral Saba, and the Ministral 3B/8B family), aiming to give organizations frontier-style capability at a substantially lower operating cost. Mistral AI frames the model as part of its broader push to accelerate enterprise adoption, with deployment paths spanning hybrid setups, on-premises installations, and in-VPC hosting, plus options for custom post-training and integration into existing toolchains.

According to Mistral AI's positioning, Mistral Medium 3 is designed to lead in professional workloads such as coding and multimodal understanding, delivering state-of-the-art results at roughly an order of magnitude less expense than larger predecessors. The model is reported to perform at or above ninety percent of Claude Sonnet 3.7 on benchmarks across the board, while also surpassing leading open models like Llama 4 Maverick and enterprise peers such as Cohere Command A on those evaluations. This combination of strong benchmark performance, coding and multimodal strengths, and flexible enterprise deployment makes the model a practical fit for teams that want near-frontier quality without the cost or infrastructure demands of the largest flagships.

Pioneermistral-mediummistral-medium

Quick Info

Powered by
Provider
Pioneer
Model key
mistral-medium
Release date
May 7, 2025
Last updated
May 7, 2025
Knowledge cutoff
2025-05
Input modalities
Output modalities
Capabilities

Cost

Input token cost
$0.40
Output token cost
$2.00

Limits

Output tokens
64,000 tokens
Context window
128,000 tokens

Transparent token rates

Compare Mistral Medium 3 pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family

Latest news about Mistral Medium 3

No articles yet. Fetch the latest news to show it here.

Videos about Mistral Medium 3

More models around Mistral Medium 3