Guide 04 of 12

Best Zero Data Retention AI

Short answer: For small and mid-sized firms, Faradex gives your team a ready-to-use Workspace on dedicated infrastructure, with conversations and files held in memory and gone when you reset. Other options include Proton Lumo for Business, which runs open-source models only, and building on a developer platform such as AWS Bedrock. For individuals, Apple Intelligence and Proton Lumo say they erase your data once they answer. Large companies can get zero data retention from OpenAI, Anthropic, Google and AWS, mostly after approval and always with exceptions. Microsoft offers a related option that switches off storage for human review.

Updated November 10, 2026

How this list works

We read each provider's own pages on November 10, 2026, and quote what they say. Each entry lists who it is for, what is retained, how you get zero data retention, a link to the source and the date it was checked. There are no ratings. In the consumer group, entries run from the least kept to the most. These are the providers' own statements; we did not audit them.

Consumer AI with zero data retention

These tools are for individuals. They range from true zero retention to up to 30 days at the model providers behind the app, and the tag on each card says which.

Apple Intelligence (Private Cloud Compute)

Zero retention, built into the design

Who it's for
People using Apple Intelligence on their Apple devices. Apple does not sell it as a business AI product.
What's retained
Nothing, by Apple's account. Apple says Private Cloud Compute "deletes the user's data after fulfilling the request, and no user data is retained in any form after the response is returned." This covers Apple's own models.
How you get ZDR
Built into the design. Apple says the encryption keys for the system's storage change on every reboot and are never kept, and there is no general-purpose logging and no remote shell.
Source
Date checked
2026-11-10

Proton Lumo

No logs while it answers; saved history is encrypted

Who it's for
Individuals, on the Free and Plus plans. The business version is listed under small and medium businesses below.
What's retained
Proton says that after Lumo answers, "the data is erased", and that Lumo doesn't store metadata such as timestamps, IP addresses or chat context. Saved chat history is stored on your device and synced to Proton's servers with zero-access encryption. Guest chats are erased at the end of each session. One exception: if you opt in, Proton keeps anonymized feedback on its Apertus 1.5 model.
How you get ZDR
It is on by default. Proton says its models "run exclusively on servers Proton controls", so no outside model provider handles your prompts.
Source
Date checked
2026-11-10

Brave Leo

Not kept by Brave, apart from a cache that lasts minutes

Who it's for
People using the AI assistant in the Brave browser. Leo Premium is an individual plan; we found no team plan.
What's retained
Brave says: "We don't store or retain prompts, responses, context, or personal data on our servers." It may cache large prompts, which are "deleted within minutes". Chat history is stored on your device. Feedback you choose to send is kept for one year.
How you get ZDR
It is on by default, under Brave's browser privacy policy. This entry covers what Brave itself keeps.
Source
Date checked
2026-11-10

Venice.ai

Depends on the mode you choose

Who it's for
Individuals. Venice Pro costs $18 a month; we found no team plan on its pricing page.
What's retained
Venice has four modes. In Anonymous mode, Venice says "you should assume the provider is storing your content." In Private mode, the default, "no prompt or response is stored", which Venice describes as "Trust-based": it relies on Venice and its partners honoring contracts. The TEE and end-to-end encrypted modes run on isolated hardware, and Venice calls end-to-end encrypted mode "Verifiably and entirely private from all parties." Chat history stays on your device.
How you get ZDR
Choose Private, TEE or end-to-end encrypted mode. TEE and end-to-end encrypted mode come with Pro.
Source
Date checked
2026-11-10

DuckDuckGo Duck.ai

Short retention: anonymized, but kept up to 30 days by model providers

Who it's for
Individuals. We found no team plan.
What's retained
Duck.ai removes metadata that identifies you, such as your IP address, before it sends your prompt to its model providers: Anthropic, Azure OpenAI, OpenAI and together.ai. Those providers must delete the data "at most within 30 days, with limited exceptions for safety and legal compliance." History stays on your device by default; optional encrypted sync is deleted if you don't access it for 18 months.
How you get ZDR
Only on some models. DuckDuckGo says "Some models have different levels of privacy protections, including Zero Data Retention."
Source
Date checked
2026-11-10

Kagi Assistant

Short retention: up to 30 days at the model providers

Who it's for
Kagi subscribers. Kagi has a Team plan: Assistant comes with standard models on the Professional tier ($10 a month per member), and premium models on the Ultimate tier ($25 a month per member).
What's retained
Kagi stores your threads, as temporary (they expire 24 hours after the last message) or permanent. Kagi's own table of model providers lists Anthropic and Bedrock at "30 days (flagged prompts: 2 years)", OpenAI at "30 days", Google at "24 hours", and some open models as "Not stored".
How you get ZDR
Not available for the frontier models. The least is kept when you use temporary threads with a model Kagi lists as "Not stored".
Source
Date checked
2026-11-10

Zero data retention AI for small and medium businesses

Most zero data retention plans are built for large companies. OpenAI and Anthropic switch it on only after a sales review, and Microsoft's closest option, modified abuse monitoring, is limited to customers "managed by a Microsoft account team or under an eligible program". Their team chat plans, ChatGPT Business and Claude Team, are not zero data retention. Small firms can buy privacy-first chat tools, or build on a developer platform such as AWS Bedrock. Faradex gives a small or mid-sized firm a ready-to-use Workspace on dedicated infrastructure, with conversations and files held in memory and gone when you reset.

Faradex

Held in memory only, on dedicated infrastructure

Who it's for
Small and mid-sized firms that want a ready-to-use AI Workspace for confidential client work.
What's retained
Files and conversations are held in memory (RAM) on your Workspace's dedicated infrastructure, with no backups. Your conversations stay in your Workspace's memory while your team works, and nowhere else. Reset it, and they're gone. Usage logs record audit details only (who, what, when and AI usage for billing), never content. Faradex staff have no access to your Workspace.
How you get ZDR
Built in for every customer. Faradex runs Amazon Bedrock in zero data retention mode, and on Bedrock, Anthropic has no access to your prompts or answers.
Model
Currently Opus via AWS Bedrock, more information is available on our website.
Security
We are SOC 2 compliant. Specifics can be found at https://faradex.ai/security.
Price
From $120 a month, no seat minimum. Team sizes are listed at https://faradex.ai/pricing.
Free trial
Claim $150 in private AI usage credit.
Source
Date checked
2026-11-10
Claim $150 in private AI usage credit

Proton Lumo for Business

No logs while it answers; open-source models only

Who it's for
Teams that want a privacy-first AI chat tool, bought per user, with admin tools.
What's retained
Proton says that after Lumo answers, "the data is erased", and that Lumo doesn't store metadata such as timestamps, IP addresses or chat context. Saved chat history is stored, with zero-access encryption so that, in Proton's words, "no one but you can read your chat history".
How you get ZDR
Self-serve. Lumo Professional costs $11.99 per user per month billed yearly, or $14.99 billed monthly, with a 14-day free trial.
Models
Open-source models only, run on servers Proton controls.
Source
Date checked
2026-11-10

These are platforms for companies that build their own AI apps. OpenAI and Anthropic switch zero data retention on after approval. Microsoft, after approval, stops storing flagged prompts for human review. Google and AWS let you configure it, with exceptions. Each one covers some features and not others.

OpenAI API

Zero retention after approval, on eligible endpoints

Who it's for
Companies building on OpenAI's API that OpenAI approves for zero data retention.
What's retained
Without zero data retention, OpenAI keeps abuse monitoring logs for all API use "for up to 30 days". With it, eligible endpoints such as /v1/chat/completions, /v1/responses and /v1/embeddings are covered. Assistants, threads, files, vector stores and batches keep data "Until deleted". An image flagged as possible child sexual abuse material is kept for manual review even with zero data retention.
How you get ZDR
Contact OpenAI's sales team. OpenAI says these controls "are subject to prior approval by OpenAI and acceptance of additional requirements." Once approved, a Data Retention tab appears under Settings, Organization, Data controls. OpenAI may make specific models ineligible for a customer, with written notice.
Source
Date checked
2026-11-10

Anthropic API

Zero retention after approval, on eligible features

Who it's for
Organizations building on Anthropic's API, including Claude Code used with commercial API keys.
What's retained
By default, Anthropic deletes API inputs and outputs "within 30 days of receipt or generation". Under zero data retention, "Anthropic does not store customer prompts or responses at rest after the API response is returned." It covers the Messages and Token Counting APIs for eligible features. It does not cover code execution, Managed Agents or Claude for Excel, or the models Anthropic calls Covered Models unless expressly authorized.
How you get ZDR
Contact Anthropic's sales team. "ZDR is enabled per organization." The Claude Team and Claude Enterprise chat apps are not eligible.
Source
Date checked
2026-11-10

Microsoft Azure OpenAI (Foundry)

Human-review storage switched off after approval

Who it's for
Customers and partners "managed by a Microsoft account team or under an eligible program".
What's retained
By default, when content is flagged, "a sample of customer's prompts and completions may be selected for review" and kept in an abuse monitoring data store. With approval for modified abuse monitoring, "the data storage and human review process described above is not performed. However, automated review may still be conducted." Stateful features, such as the Responses API, Assistants threads and stored completions, store data in your own Azure tenant.
How you get ZDR
Apply for modified abuse monitoring with Microsoft's form. After approval, check that "ContentLogging" shows "false".
Source
Date checked
2026-11-10

Google Vertex AI (now Gemini Enterprise Agent Platform)

Zero retention once you switch several features off

Who it's for
Google Cloud customers building with Google's models, or with OpenAI models on Google's platform.
What's retained
Under zero data retention, Google says it processes requests "without writing prompts, file inputs, or responses to persistent disk." It still uses an in-memory cache that is cleared within 24 hours, and it records operational and billing metadata such as token counts. Google says prompt logging for abuse monitoring may apply under its Cloud terms. Grounding with Google Search keeps logs for up to three days, and Google says there is no way to disable that. Grounding with Google Maps keeps them for 30 days. Deep Research keeps them for seven days, which cannot be disabled.
How you get ZDR
For Google's models, switch off caching and request-response logging yourself, set store=false on the Interactions API, and don't use Grounding with Google Search or Maps. For OpenAI models on Google, zero data retention is "available ... on an approved project basis, but is disabled by default."
Source
Date checked
2026-11-10

AWS Bedrock

Zero retention by account setting, with exceptions for some models

Who it's for
Any AWS account. It is a developer platform, so someone has to build or host the app your team uses.
What's retained
With the setting at none, "No request or response data is written to durable storage by AWS or shared with the model provider." AWS names some models whose traffic is kept for up to 30 days for automated abuse detection; for OpenAI models other than the gpt-oss family, flagged traffic is kept for up to 30 days. AWS may store and review flagged content only to check for child sexual abuse material. Model providers "don't have access to Amazon Bedrock logs or to customer prompts and completions."
How you get ZDR
Self-serve. Set data_retention_mode to none for your account or project, in each Region; new accounts start at inherit. If a model requires retention, "Amazon Bedrock will block the request and return an error." For OpenAI models, "Eligible customers may request full ZDR through their AWS account team." AWS also says customers eligible for its "Enterprise Frontier Safeguards program will receive ZDR through June 30, 2027" on the models whose traffic it keeps for 30 days.
Source
Date checked
2026-11-10

About Faradex. Faradex is a private AI workspace for small and mid-sized firms that handle confidential client work. Each firm gets its own Workspace on dedicated infrastructure in the United States. Files and conversations are held in memory and are gone when the Workspace is reset. AI model: Currently Opus via AWS Bedrock, more information is available on our website. AWS does not store prompts or responses, and model providers have no access to them. Pricing: from $120 a month, no seat minimum; see faradex.ai/pricing.