Privacy Policy

ocoolai ("we", "us") operates ocoolai (one.ocoolai.com) (the "Service"). This policy explains, in plain language, what information we collect, why we collect it, how we use and protect it, who we share it with, how long we keep it, and what rights you have over your data.

Please read this policy in full before using the Service. If you do not agree with any part of it, please stop using the Service.

1. Scope

This policy covers all interactions with us through the web interface, the application programming interface (API), and every feature module of the Service, including AI Chat, AI Studio, deep research, the model catalogue, and account and billing functions.

This policy does not cover:

• the independent processing of your inputs and outputs by the third-party model providers you choose to route requests to, which is governed by their own privacy policies;
• third-party tools, MCP servers, or external search services that you connect yourself;
• third-party websites you navigate to from the Service.

2. Information We Collect

2.1 Information you provide

2.2 Information collected automatically

2.3 Information from third parties

When you sign in or link an account through a third-party provider, we receive the basic identifiers that provider exposes to us (typically user ID, username, and avatar, depending on the scope you authorise). We never receive your password on that platform, and we do not read data beyond the authorised scope.

3. How Each Module Handles Your Data (Key Section)

Different modules treat your inputs in materially different ways. Please read this section carefully.

3.1 API relay: we do not retain request or response bodies

When you call our model endpoints with an API key:

• we do not store your request body (prompts, messages, images, files) or the model's response body;
• request content is held in memory only for as long as forwarding requires, and is released immediately afterwards;
• we record request metadata only, as listed in section 2.2, for billing, usage reporting, and troubleshooting;
• the fields visible on your "Logs" page are the complete set of what we retain, and they contain no conversation content.

Important: this commitment binds us only. Once forwarded, the upstream model provider decides independently whether and for how long to retain the request under its own policy. Practices vary widely — some providers retain data for several days for abuse monitoring, while others offer zero-retention endpoints. Verify the target provider's data policy before sending sensitive material.

3.2 AI Chat (local mode, the default)

With cloud sync switched off:

• your thread list, message content, model settings, and long-term memory are stored only in your browser's local storage (IndexedDB / localStorage);
• this data is never uploaded to our servers and we cannot read it;
• clearing browser data, switching browsers, or changing devices will make it unrecoverable.

In local mode, any custom API key and custom endpoint you enter also stay in your browser. Cloud sync strips these sensitive fields before uploading, so they never leave your device as part of synced data.

3.3 AI Chat cloud sync (optional, opt-in)

Cloud sync is an optional feature you enable yourself in order to synchronise conversations across devices. Once enabled, your thread titles, message bodies, model configuration, and attachments are uploaded to and stored in our server databases and object storage in order to deliver that synchronisation. Unsent composer drafts are not part of cloud sync and stay on your device.

• Synced content is stored in structured form. It is encrypted in transit with HTTPS/TLS and protected by encryption at rest at the storage layer, but it is not end-to-end encrypted at field level, which means we are technically able to access it.
• We access that content only: (a) to deliver the synchronisation, retrieval, and export functions you request; (b) to investigate a fault you have reported to us; or (c) to comply with a lawful, binding demand from a competent judicial or administrative authority.
• We do not use that content to train any model, and we do not sell, rent, or otherwise trade it.
• You may switch cloud sync off at any time, or delete synced threads. Deletion is described in section 7.

If you handle highly sensitive information, we recommend leaving cloud sync off, using local mode only, and keeping your own backups.

3.4 AI Studio

When you generate images, video, or audio with AI Studio:

• your prompts and generation parameters are stored so that you can review your creation history, reuse parameters, and trace problems;
• links to the generated results and their metadata (model, dimensions, status, file size) are stored in your asset library;
• some outputs are distributed through a temporary hosting service, where files are cleared automatically after approximately 7 days;
• you may delete individual assets or entire creation topics at any time.

3.5 Shared conversations

When you create a share link:

• we save a content snapshot of that conversation;
• anyone holding the link, including signed-out visitors, can view the snapshot, and search engines may index it;
• the snapshot expires at the time you set, and you can revoke sharing at any point.

Please confirm the conversation contains no sensitive personal data, credentials, or trade secrets before sharing.

4. How We Use This Information

We process your information only for the following purposes:

1. Delivering the Service — authenticating your account, routing and forwarding model calls, returning generated results, and synchronising and displaying your data.
2. Billing and settlement — calculating usage and charges from request metadata, and processing top-ups and invoices.
3. Security and abuse prevention — detecting anomalous sign-ins, preventing credential stuffing and fraudulent usage, blocking unlawful use, and maintaining availability.
4. Troubleshooting and improvement — diagnosing errors and improving performance and stability, using aggregated and anonymised statistics rather than individual conversation content.
5. Essential notices — sending you notifications about account security, balance, and service changes.
6. Legal obligations — meeting obligations imposed on us by applicable law.

Where the General Data Protection Regulation (GDPR) applies, our legal bases are: performance of our contract with you (Art. 6(1)(b)); our legitimate interests (Art. 6(1)(f), for example security); your consent (Art. 6(1)(a), for example cloud sync and optional analytics); and compliance with legal obligations (Art. 6(1)(c)).

5. What We Will Never Do

• We do not train on your inputs or outputs. Your conversations, prompts, and generated results are never used to train, fine-tune, or evaluate any model.
• We do not sell your personal information. We have never sold, rented, or traded your personal information for money or other valuable consideration. Under the CCPA/CPRA, we neither "sell" nor "share" personal information for cross-context behavioural advertising.
• We do not use conversation content for advertising or profiling.
• We do not access your synced conversations beyond the purposes stated in this policy.
• We do not expand our collection silently. If we need to process your information for a new purpose, we will tell you and, where required, obtain your consent.

6. Sharing and Third Parties

We share your information only where necessary:

6.1 Upstream model providers

To fulfil a model call you initiate, your input must be transmitted to the provider you selected. This transfer is essential to the service, and the provider applies its own privacy policy and retention rules to what it receives. You control which provider receives your data through your choice of model.

6.2 Infrastructure providers

We use object storage for attachments and generated files, and cloud servers and managed databases to run the Service. These providers process data only as needed to serve us, are bound by contractual confidentiality obligations, and may not use your data for their own purposes.

6.3 Identity providers

When you use social sign-in, your authentication request is exchanged with the relevant identity provider.

6.4 Site analytics (optional, depending on site configuration)

This site may have web analytics enabled to measure traffic and page performance. Such tools collect page-level information and never touch your conversation content. Where your jurisdiction requires consent for analytics, we will obtain it before enabling them.

6.5 Legal requirements

We may be compelled to disclose information in response to a lawful demand from a competent judicial, law-enforcement, or regulatory authority. To the extent permitted by law, we review such demands for validity and necessity and notify you in advance where we are allowed to.

6.6 Corporate transactions

In a merger, acquisition, or asset transfer, your information may be transferred as part of those assets. We will notify you before any such transfer, and the successor must remain bound by protections no weaker than this policy.

7. How Long We Keep Data

After these periods we delete or anonymise the data. Aggregated, anonymised statistics (such as how many times a model was called in a month) no longer relate to you and may be retained.

8. Cookies and Local Storage

We use a small number of necessary cookies to maintain your session and preferences:

We use no third-party advertising cookies and perform no cross-site tracking.

The Service also stores conversation data, interface settings, and caches in your browser's local storage (IndexedDB / localStorage). You may clear these at any time through your browser, which will permanently remove any unsynced local data.

9. Security

We apply technical and organisational measures proportionate to the risk, including:

• HTTPS/TLS enforced across the entire site;
• passwords stored using an irreversible hashing algorithm, so we cannot recover your plaintext password;
• tokens and third-party credentials encrypted at rest in the database;
• session credentials issued as high-entropy random identifiers, marked HttpOnly to resist script-based theft, and subject to an absolute lifetime cap;
• strict database privilege separation, with access to production data on a least-privilege basis;
• input validation, parameterised queries, and rate limiting at the API layer.

Please understand that no method of internet transmission or storage is completely secure. We cannot guarantee absolute security. Keep your password and API keys safe, never enter credentials on unofficial pages, and never commit API keys to public code repositories. If you notice anything unusual about your account, change your password immediately and contact us.

10. Your Rights

Wherever you are located, you may exercise the following rights:

• Right to be informed — to understand how we process your information, as set out in this policy;
• Right of access and portability of a copy — to see and obtain a copy of the information we hold about you;
• Right to rectification — to correct inaccurate or incomplete information; most fields can be edited yourself in your profile;
• Right to erasure — to delete your conversations, assets, and share links; to delete your account and all associated data, contact us using the details below;
• Right to data portability — to export your conversation data in a structured, commonly used, machine-readable format, which the Service supports natively;
• Right to withdraw consent — to turn off consent-based features such as cloud sync at any time; withdrawal does not affect the lawfulness of processing carried out beforehand;
• Right to restriction and objection — to restrict or object to particular processing where applicable law provides for it;
• Right against discrimination — we will not degrade your service or raise your price because you exercised these rights.

Most actions can be completed yourself in the product (settings, asset library, share management). For anything else, contact us as described in section 14. We may need to verify your identity first to protect your account. We will respond within a reasonable period, normally no more than 30 days.

If you believe our processing infringes your rights, you may lodge a complaint with your local data protection authority.

11. International Transfers

Our infrastructure and the model providers you select may be located outside your country or region. By using the Service, your data may be transferred to and processed in those locations.

For users in the European Economic Area, the United Kingdom, or Switzerland, transfers to countries without an adequacy decision rely on Standard Contractual Clauses or another transfer mechanism recognised by applicable law, together with proportionate supplementary safeguards.

12. Minors

The Service is not directed at anyone under 18. We do not knowingly collect personal information from minors.

If you are a guardian and discover that a minor has used the Service without your consent, please contact us and we will delete the associated information and close the account after verification.

In jurisdictions whose law permits users aged 13 to 18 to use comparable services with guardian consent, such use requires that guardian's explicit prior consent.

13. Region-Specific Disclosures

13.1 European Economic Area, United Kingdom, and Switzerland (GDPR / UK GDPR)

• We act as the "data controller" for the processing described in this policy.
• Our legal bases are set out in section 4.
• You have all data subject rights under Articles 15 to 22 GDPR; see section 10 for how to exercise them.
• You may lodge a complaint with the supervisory authority in your member state.
• We do not carry out solely automated decision-making producing legal or similarly significant effects.

13.2 California residents (CCPA / CPRA)

• The categories of personal information we collected in the past 12 months are in section 2, our purposes in section 4, and the categories of recipients in section 6.
• We have not sold personal information and have not shared it for cross-context behavioural advertising, including that of consumers under 16.
• You have the rights to know, delete, correct, limit the use of sensitive personal information, and not be discriminated against for exercising them.
• You may use an authorised agent, from whom we may require written proof of authorisation.

13.3 Other jurisdictions

Where your local data protection law grants rights not expressly listed here, we will honour them to the extent that law applies. Each site operated on this platform must comply with the laws of its place of operation and of its target users, including any sector-specific rules governing generative AI services.

14. Changes and Contact

We may revise this policy to reflect product changes, legal developments, or security needs. For material changes — such as broadening what we collect, changing categories of recipients, or altering retention — we will notify you through an in-product announcement, email, or a prominent notice at sign-in, and seek fresh consent where required. Revisions take effect on publication, and continued use after that date constitutes acceptance.

For questions or to exercise your rights, please reach us through the support or ticket channel provided on this site.

Operator: ocoolai