Mistral now trains on user input by default, except on enterprise tier
Heat trend
Collecting trend data
The percentage is based on available heat signal, not comment count or independent people.
An official release brings Mistral model updates — worth tracking for capability changes, ecosystem impact, and follow-up.
- Basis
- Running about 2.3× the median of this source's recent listed items
- Metric comparison
- 152 vs median 65.5 (20 baseline samples)
- Detected
- 09/02, 16:01
Mistral now trains its models on user input and output data by default, with an exception for its enterprise tier. Users' conversations, documents, and other content may be included in training programs. The opt-out process varies depending on the service or platform used, including Vibe (via Admin panel or mobile apps on iOS and Android) and Mistral Studio and API services (via Admin panel). Related articles address data retention, AI model training, conversation privacy, GDPR rights, and data handling for the Memories feature.
Updated this week
In certain cases, your input and output data (such as conversations, documents, and other user-provided content) may be included in Mistral’s model training programs.
The opt-out process depends on the service or platform used, as described below.
Opt out of Vibe data training (via the Admin panel)
You can opt out of training through the Admin panel by following the procedure below:
Opt out of Vibe data training on the mobile application (iOS and Android)
On both iOS and Android, the opt-out procedure is as follows:
Opt out of Mistral Studio and API data training (via the Admin panel)
You may opt out of data training for Mistral Studio and related API services by following the steps below:
Related Articles
- Can I activate Zero Data Retention (ZDR)?
- Do you use my user data to train your Artificial Intelligence models?
- Can other people view my conversations?
- How can I exercise my GDPR rights?
- How do you handle my data when using the Memories feature?