Key Facts
What is each party's total aggregate liability capped at?
Each party's total aggregate liability arising out of or relating to the Terms or the Mistral AI Products is capped at the total amounts paid by Customer to Mistral AI in the twelve months preceding the event(s) giving rise to the claim.
Is liability capped at amounts paid in the twelve months preceding the event?
Each party's total aggregate liability arising out of or relating to the Terms or the Mistral AI Products is capped at the total amounts paid by Customer to Mistral AI in the twelve months preceding the event(s) giving rise to the claim.
Are amounts paid by the Customer non-refundable?
Mistral AI treats all amounts paid by the Customer as non-refundable, non-cancellable, and non-creditable.
Are amounts non-cancellable and non-creditable?
Mistral AI treats all amounts paid by the Customer as non-refundable, non-cancellable, and non-creditable.
Is the Customer required to indemnify Mistral AI?
The Customer is required to indemnify, defend, and hold harmless Mistral AI and its affiliates and licensors against third-party liabilities, damages, and costs arising from the Customer's use of the Mistral AI Products in violation of the Terms.
Must the Customer indemnify against third-party liabilities arising from violation of the Terms?
The Customer is required to indemnify, defend, and hold harmless Mistral AI and its affiliates and licensors against third-party liabilities, damages, and costs arising from the Customer's use of the Mistral AI Products in violation of the Terms.
What must the Customer comply with?
Mistral AI requires the Customer to comply with all applicable export control, economic sanctions, and trade compliance laws and regulations, including those of the European Union, the United States, and Singapore.
Does Mistral AI require the Customer to comply with export control, economic sanctions, and trade compliance laws and regulations?
Mistral AI requires the Customer to comply with all applicable export control, economic sanctions, and trade compliance laws and regulations, including those of the European Union, the United States, and Singapore.
Who is solely responsible for activities conducted through the Customer Account?
Mistral AI holds the Customer solely responsible for any activities conducted through or under its Customer Account (including End User Accounts), in connection with Customer Infrastructure, or in connection with Customer Offerings.
Is the Customer solely responsible for activities in connection with Customer Infrastructure?
Mistral AI holds the Customer solely responsible for any activities conducted through or under its Customer Account (including End User Accounts), in connection with Customer Infrastructure, or in connection with Customer Offerings.
Summary
This document sets out the rules for businesses that pay to use Mistral AI's products. Every payment is final—no refunds, cancellations, or credits—and if something goes wrong, the most either side can recover is the amount the Customer paid Mistral AI in the prior twelve months. By default, data and outputs generated through Labs Models may be used by Mistral AI to train its AI, but the Customer can prevent this by activating zero data retention.
Analysis
These commercial terms establish the contractual framework governing Customer access to and use of Mistral AI Products, setting out mutual obligations and financial allocations between the parties. Liability exposure is capped bilaterally at fees paid in the twelve months preceding a claim, consequential and indirect damages are excluded for both parties and their affiliates, and all amounts paid by the Customer are non-refundable and non-creditable. Mistral AI provides its products, documentation, and configuration recommendations on an as-is basis with no warranties, while the Customer bears sole responsibility for all activity conducted through its account, infrastructure, and customer-facing offerings. Data use rights are bifurcated: Mistral AI holds a worldwide royalty-free license for operational purposes excluding model training, but may use Customer Data and Outputs from Labs Models for model training unless the Customer activates zero data retention. Each party carries a defined indemnification obligation—the Customer indemnifies Mistral AI against third-party claims arising from Terms violations, and Mistral AI indemnifies the Customer against third-party intellectual property infringement claims.
What this means for you
As an enterprise customer, you are solely responsible for everything that happens through your account and any products you build on top of Mistral AI. All fees you pay are non-refundable under any circumstance stated in these terms, and your financial recovery in any dispute is capped at what you paid in the twelve months before the claim arose. If you use Labs Models, your data and outputs are used by Mistral AI for model training by default—you can prevent this by activating the zero data retention setting. If you provide Feedback to Mistral AI, the associated data and outputs become freely usable by Mistral AI without any compensation owed to you.
2 important changes detected
3 versions captured · Last updated: August 2026
What changed
Mistral AI revised its Commercial Terms effective August 5, 2026, with the main operational changes being: the company expanded the definition of Customer Data to include 'access to data' and 'coding environment'; restructured its AI training opt-out framework to be clearer about which products use training by default and which require opt-in; consolidated references to 'Labs Models' and newly defined 'Preview Models' as a single category ('Labs or Preview Models') with the same data-training rules; and removed a carve-out from its liability cap that previously protected customers against breaches of confidentiality obligations, narrowing the exceptions to liability limitations.
Why this matters
The updated terms redefine when Mistral AI will use customer data and outputs for AI model training. Previously, the terms distinguished between free tiers (where training was opt-out) and Labs Models (where training always occurred). The revised language consolidates experimental models as 'Labs or Preview Models' and explicitly states that training opt-out preferences applied to other products do not carry over to these experimental models. Additionally, the terms now specify that customer data includes 'access to data' and 'coding environment,' expanding what qualifies as trainable material. The liability cap was narrowed to remove the carve-out protecting confidentiality breach claims, meaning liability limits apply more broadly. You can review which Mistral products are set to opt-in versus opt-out by default and avoid Labs or Preview Models if you do not want your data used for training.
View full change record →
What changed
Mistral AI updated its Commercial Terms on May 29, 2026, replacing references to 'Le Chat Pro' and 'Le Chat Teams' with 'Vibe Pro' and 'Vibe Teams' throughout the document. This reflects a product naming change. The substantive exception to the prohibition on using customer data for model training remains the same: Mistral AI will not use customer data for training except under free subscriptions, paid tiers, feedback scenarios, automated moderation flags, Order Form exceptions, or Labs Models usage. The operational framework for data use authorization is unchanged.
Why this matters
The updated terms reflect a product naming change from Le Chat to Vibe. The substantive policy governing when Mistral AI may use customer data for model training remains unchanged. Under both versions, Mistral AI does not use customer data for training except in specified scenarios: free subscriptions, paid tiers where the customer has not opted out, user feedback, flagged content for moderation, exceptions in an Order Form, or use of Labs Models. No material change to your data usage rights or restrictions occurs.
View full change record →
Archival ProvenanceSource & Archival Record
Last Captured
August 8, 2026 00:50 UTC
Capture Method
Automated scheduled archival capture
Document ID
CA-D-000769
Version ID
CA-V-005624
SHA-256
ebca926d5afbb123bdf69562e094d48077e63b966030f0612cc42f8282c654b5
✓ Snapshot stored
✓ Text extracted
✓ Change verified
✓ Hash verified