The page states that this Usage Policy is effective June 11, 2026. The source shows no other version, successor or future-dated text.
Supporting quotations
Effective: June 11, 2026
Legal Rights & Agreement Rules · Context
Policy covers Mistral AI products on its platform, their users, content and activity
The Usage Policy applies to the Mistral AI products accessible on Mistral's platform or applications (the source names Vibe and Mistral AI Studio as examples), to all users of those products, whether individuals, organizations or businesses, to all content generated, shared or accessed through them, and to all interactions and activities on or through them.
Supporting quotations
Our usage policy (the “Usage Policy”) applies to:
All the Mistral AI products accessible on our platform or applications, including Vibe and Mistral AI Studio
All users of our Mistral AI Products accessible on our platform, including individuals, organizations, and businesses
All content generated, shared, or accessed through our Mistral AI Products accessible on our platform
All interactions and activities conducted on or through the Mistral AI Products accessible on our platform.
Qualifications & exceptions
The scope is tied to products accessible on Mistral's platform or applications; deployments outside that platform are excluded elsewhere in the policy.
Not stated here
The source names only Vibe and Mistral AI Studio as examples ('including'); it does not list every product, so whether a particular other product is covered is not stated in this document.
Cross-references
has exception (model-proposed): Policy does not apply to customer-hosted, partner-hosted or open-source offerings — The scope list covers Mistral AI products accessible on its platform or applications; the exclusion removes products deployed on a customer's or partner's infrastructure and open-source models and products from the policy.
Legal Rights & Agreement Rules · Restriction or exception
Policy does not apply to customer-hosted, partner-hosted or open-source offerings
The Usage Policy does not apply to Mistral AI products deployed on a customer's infrastructure, to products deployed on the infrastructure of Mistral's partners, or to Mistral's open-source AI models and products.
Supporting quotations
This Usage Policy does not apply to Mistral AI Products deployed on a customer’s infrastructure, on the infrastructure of our partners, or to our open-source AI models and products.
Not stated here
This document does not say which terms, if any, govern the excluded deployments.
Cross-references
exception to (model-proposed): Policy covers Mistral AI products on its platform, their users, content and activity — The scope list covers Mistral AI products accessible on its platform or applications; the exclusion removes products deployed on a customer's or partner's infrastructure and open-source models and products from the policy.
Mistral states principles of neutrality, user empowerment and transparency
Mistral says the following principles guide how it builds and offers its services: technology should be as neutral as possible; it aims to minimize the risk of abuse while trusting that users should be empowered to define and use available controls adapted to them and their priorities; and it aims to be transparent about its AI models, policies and enforcement while giving everyone access to generative AI. These are statements of aims and belief, not specific rules or commitments about particular conduct.
Supporting quotations
the following principles guide how we build and offer our services:
Be neutral: we believe technology should be as neutral as possible.
we believe in minimizing the risk of abuse while trusting that you should be empowered to define and use available and robust controls adapted to you and your priorities.
we believe in building services that provide everyone the right to access generative AI and in being transparent about our artificial intelligence models, policies, and enforcement.
Qualifications & exceptions
Phrased as beliefs and guiding aims ('we believe').
Not stated here
The source does not say which user-adjustable 'controls' exist or how a user defines them.
The source does not say what 'neutral' means in practice or how it interacts with the prohibited uses listed later.
Legal Rights & Agreement Rules · Restriction or exception
Using Mistral AI products for illegal activity is prohibited and may be reported to law enforcement
Users must comply with applicable laws when using the Mistral AI products. They must not use the products to engage in, encourage, promote, offer or solicit illegal activity, and any such use may be reported to law enforcement. This prohibition, like the rest of the policy, does not cover products deployed on a customer's or partner's infrastructure or Mistral's open-source models and products.
Supporting quotations
you must comply with applicable laws. As such, you must not use the Mistral AI Products to engage in, encourage, promote, offer, or solicit illegal activity, and any such use may be reported to law enforcement.
This Usage Policy does not apply to Mistral AI Products deployed on a customer’s infrastructure, on the infrastructure of our partners, or to our open-source AI models and products.
Qualifications & exceptions
Reporting is stated as 'may', not 'will'.
Does not apply to Mistral AI products deployed on a customer's or partner's infrastructure, or to open-source models and products.
Not stated here
The source does not say which jurisdiction's laws count as 'applicable laws'.
As examples of what users must not use the Mistral AI products to generate, the policy lists content that promotes, facilitates or gives instructions for drug trafficking or distribution, or information on obtaining or using illegal substances; content that infringes or violates a third party's rights, including intellectual property rights; content related to activities with a high risk of physical harm, such as weapons development, military and warfare, or management or operation of critical infrastructure in energy, transportation and water; and content that enables or exploits individuals, including through human trafficking or sexual services. The list sits under the 'Illegal activities' paragraph, and the physical-harm item is worded broadly: it does not say whether all such content is barred or only content that is itself illegal.
Supporting quotations
Content that promotes, facilitates, or provides instructions for drug trafficking or distribution, or information on obtaining or using illegal substances
Content that infringes, misappropriates, or otherwise violates any third party’s rights, including intellectual property rights
Content related to activities with high-risk of physical harm, such as weapons development, military and warfare, management, or operation of critical infrastructure in energy, transportation, and water
Content that enables or exploits individuals, including through human trafficking or sexual services.
Qualifications & exceptions
Introduced as examples ('For instance') under the illegal-activities paragraph.
Not stated here
Whether the physical-harm item (weapons development, military and warfare, critical infrastructure) bars such content in every case or only where it is illegal is not stated.
The source's wording 'management, or operation of critical infrastructure' is run together and does not say whether management and operation are separate activities.
Legal Rights & Agreement Rules · Company commitment
Child sexual abuse material: zero tolerance, reporting and account termination
Mistral states a zero tolerance policy on child sexual abuse material (CSAM). Generating, or attempting to generate, CSAM on the Mistral AI products is strictly prohibited. Mistral says it will report all actual or suspected CSAM to the relevant law enforcement authorities and will terminate the user's account.
Supporting quotations
Mistral AI has a zero tolerance policy regarding CSAM. Any generation or attempt to generate CSAM on our Mistral AI Products is strictly prohibited. We will report all actual or suspected CSAM to the relevant law enforcement authorities and will terminate your account.
Qualifications & exceptions
Covers attempts to generate, and suspected as well as actual CSAM, for reporting.
Not stated here
The source does not say who decides that CSAM is 'suspected' or whether the user is told of a report.
Cross-references
contrast (model-proposed): Violations may lead to suspension or termination and reports to authorities — The general enforcement paragraph says violations 'may' lead to suspension or termination and reporting 'where appropriate', while the CSAM paragraph says Mistral 'will' report and 'will terminate' the account. The source does not say whether other violations can also lead to automatic termination.
Legal Rights & Agreement Rules · Restriction or exception
Generating intimate images of a person requires every person's explicit consent
Users must not use the Mistral AI products to generate intimate images of any person without the explicit consent of all individuals involved.
Supporting quotations
You shall not use the Mistral AI Products to generate intimate images of any person without the explicit consent of all individuals involved.
Not stated here
The source does not define 'intimate images' or say how consent is shown.
Legal Rights & Agreement Rules · Restriction or exception
Content promoting hate, discrimination or harassment is prohibited
Users must not use the Mistral AI products to promote hate, discrimination or harassment, including generating content that promotes, incites, glorifies or celebrates: hate or discrimination based on race, gender, ethnicity, religion, nationality, sexual orientation, disability status or caste; revisionism or denialism, particularly regarding historically documented events such as the Holocaust or other genocides; harassing language toward any target; or the suffering or humiliation of others.
Supporting quotations
You shall not use our Mistral AI Products to promote hate, discrimination, or harassment, including generating any content that promotes, incites, glorifies, or celebrates:
Hate or discrimination based on an individual’s race, gender, ethnicity, religion, nationality, sexual orientation, disability status, or caste
Engaging in revisionism or denialism, particularly regarding historically documented events such as the Holocaust or other genocides
Harassing language towards any target
The suffering or humiliation of others.
Qualifications & exceptions
The list is introduced with 'including', so it does not say it is exhaustive.
Not stated here
The source does not define 'revisionism or denialism' beyond the historical examples it gives.
Legal Rights & Agreement Rules · Restriction or exception
Promoting, inciting, threatening or glorifying violence is not permitted
Users may not use the Mistral AI products to conduct or generate any activity or content that promotes, incites, threatens or glorifies violence. The policy gives threats of physical harm and glorification of violent acts or terrorism as examples.
Supporting quotations
Use of the Mistral AI Products to conduct or generate any activity or content that promotes, incites, threatens, or glorifies violence is not permitted. This includes, for instance, threats of physical harm or glorification of violent acts or terrorism.
Not stated here
The source does not say whether fictional or depicted violence is covered.
Legal Rights & Agreement Rules · Restriction or exception
Content promoting or instructing self-harm or suicide is prohibited
Users must not use the Mistral AI products to generate content that promotes, encourages or provides instructions for self-harm or suicide. Examples given are graphic depictions of self-harm, detailed descriptions of self-harm methods, and encouragement of others to engage in self-harm or suicidal behavior.
Supporting quotations
You shall not use the Mistral AI Products to generate content that promotes, encourages, or provides instructions for self-harm or suicide.
Graphic depictions of self-harm
Detailed descriptions of self-harm methods
Encouragement of others to engage in self-harm or suicidal behavior.
Qualifications & exceptions
Examples are introduced with 'for instance'.
Legal Rights & Agreement Rules · Restriction or exception
Fraud and scams, including phishing and impersonation, are prohibited
Using the Mistral AI products to engage in any form of fraudulent activity or scams is prohibited. The policy's examples are phishing attempts, financial scams, impersonation, and misrepresentation of facts and data.
Supporting quotations
Using the Mistral AI Products to engage in any form of fraudulent activity or scams is prohibited.
Phishing attempts
Financial scams
Impersonation
Misrepresentation of facts and data.
Qualifications & exceptions
Examples are introduced with 'for instance'.
Not stated here
The source does not say whether any 'misrepresentation of facts and data' is barred or only deceptive use.
Legal Rights & Agreement Rules · Restriction or exception
Deliberately misleading or deceptive content is prohibited
Users must not use the Mistral AI products to generate content that is deliberately misleading, false or intended to deceive others. Examples given are spreading false information about current events, health or science, or that undermines the integrity of a civic or political process; promoting conspiracy theories that can cause harm; and misinformation that targets protected groups. The lead-in speaks of deliberate or deceptive content, but the second and third examples do not repeat an intent requirement, and the source does not say how the two fit together.
Supporting quotations
You shall not use the Mistral AI Products to generate content that is deliberately misleading, false, or intended to deceive others.
Spreading false information about current events, health, science, or that undermines the integrity of a civic or political process
Promoting conspiracy theories that can cause harm
Misinformation that targets protected groups.
Qualifications & exceptions
The lead-in uses 'deliberately misleading, false, or intended to deceive'; its grammar does not say whether 'deliberately' also governs 'false'.
Not stated here
Whether false content generated without intent to deceive falls inside the prohibition is not stated.
Who decides what is a conspiracy theory 'that can cause harm' is not stated.
Legal Rights & Agreement Rules · Restriction or exception
Generating content that invades others' privacy, including likeness or voice without consent, is prohibited
Generating content that invades or violates the privacy of others is prohibited. As an example, the policy names using someone else's likeness or voice to generate outputs or to impersonate them without their prior consent.
Supporting quotations
Generating content that invades or violates the privacy of others is prohibited. This includes, for instance, using someone else’s likeness or voice to generate outputs or impersonate them, without their prior consent.
Not stated here
The source does not say how 'privacy' is defined or whether the consent exception applies beyond the likeness and voice example.
Legal Rights & Agreement Rules · Restriction or exception
Providing professional advice without proper qualification is prohibited
Users shall not use the Mistral AI products to provide professional advice without proper qualification. The policy's examples are offering investment advice, financial planning or any form of financial guidance; providing legal counsel, interpretations of the law or recommendations on legal actions; and offering medical diagnoses, treatment suggestions or any form of health-related guidance. The wording is about using the products 'to provide' advice. It does not say who must hold the qualification (the user or the product), and the examples are not themselves limited by the qualification wording.
Supporting quotations
You shall not use our Mistral AI Products to provide professional advice without proper qualification.
Offering investment advice, financial planning, or any form of financial guidance;
Providing legal counsel, interpretations of the law, or recommendations on legal actions;
Offering medical diagnoses, treatment suggestions, or any form of health-related guidance.
Qualifications & exceptions
The rule is qualified by 'without proper qualification'; 'proper qualification' is not defined.
Not stated here
The source does not say what counts as 'proper qualification' or whose qualification is meant.
The source does not say whether a user who only reads the product's general information on these topics, without passing it on as advice, is affected.
Legal Rights & Agreement Rules · Permission or license
Mistral may update the policy on its website; users are told to check it
Mistral reserves the right to update this Usage Policy from time to time on its website, and invites users to consult it frequently to stay informed of changes. The source states no advance notice, no individual notification and no effective-date mechanism for changes. The same paragraph also reserves the right to update the 'Terms of Use' at any time, a document name that differs from the 'Terms of Service' named in the enforcement paragraph.
Supporting quotations
We reserve the right to update this Usage Policy from time to time on our website.
We reserve the right to update the Terms of Use at any time. We therefore invite you to consult our Usage Policy frequently to stay informed of any changes.
Not stated here
The source does not say whether users are notified of changes or how continued use is treated.
It is not stated whether 'Terms of Use' and 'Terms of Service' are the same document; neither was supplied.