Private and open-source alternatives to Perplexity
12 AI assistants rated against the same public privacy criteria as Perplexity, with our picks first.
Why look for an alternative?
- Open source: no. Closed source.
- No trackers or telemetry: no. The Android app includes Google Firebase Analytics, Crashlytics and Singular.
- No ads or data sales: partial. Sponsored follow-up questions and ads can appear next to answers. Perplexity states queries are not sent to advertisers.
- Independent audit: no. No independent audit of the consumer service is published.
- Transparency report: no. No transparency report or government request policy is published.
- Tells users about requests: no. No published policy on notifying users about data requests.
- Security headers: partial. Grade B+ (80/100+)
- No training on your data: partial. Searches are used to improve the service by default. The AI Data Usage setting turns this off.
- Runs locally: no. Hosted only.
- Limited chat retention: partial. Search history is kept while the account is active and can be deleted. Account data is removed within 30 days of account deletion.
- No account needed: partial. Search works without an account, with fewer features.
See the full Perplexity rating.
The alternatives
Jan Grade B
Open-source desktop app for running large language models offline on a local computer, with optional connections to remote AI APIs.
Jan scores 89 out of 100 (grade B) on the AI assistants criteria. It meets 7 of 8 criteria: open source, no trackers or telemetry, no ads or data sales, no training on your data, runs locally, limited chat retention and no account needed.
llama.cpp Grade B
Open-source C/C++ library and command-line tools for running large language model inference locally on CPUs and GPUs, including a built-in web server and chat interface.
llama.cpp scores 89 out of 100 (grade B) on the AI assistants criteria. It meets 7 of 8 criteria: open source, no trackers or telemetry, no ads or data sales, no training on your data, runs locally, limited chat retention and no account needed.
Ollama Grade B
Open-source tool for downloading and running large language models on a local computer through a command line, API and desktop app. Optional cloud-hosted models are also offered.
Ollama scores 81 out of 100 (grade B) on the AI assistants criteria. It meets 6 of 8 criteria: open source, no ads or data sales, no training on your data, runs locally, limited chat retention and no account needed.
Lumo Grade C
AI assistant from Proton that runs open-weight models on Proton's servers, with no logs of chats and zero-access encrypted chat history.
Based in Switzerland. Lumo scores 73 out of 100 (grade C) on the AI assistants criteria. It meets 7 of 12 criteria: no ads or data sales, transparency report, tells users about requests, TLS configuration, security headers, no training on your data and limited chat retention.
GPT4All Grade C
Open-source desktop app from Nomic for running large language models locally, including chatting with local documents.
GPT4All scores 72 out of 100 (grade C) on the AI assistants criteria. It meets 6 of 8 criteria: open source, no ads or data sales, no training on your data, runs locally, limited chat retention and no account needed.
LM Studio Grade C
Desktop app for finding, downloading and running large language models locally, with a chat interface and a local API server.
LM Studio scores 72 out of 100 (grade C) on the AI assistants criteria. It meets 6 of 8 criteria: no trackers or telemetry, no ads or data sales, no training on your data, runs locally, limited chat retention and no account needed.
Open WebUI Grade C
Self-hosted web interface for large language models that works with Ollama and OpenAI-compatible APIs, with multi-user support.
Open WebUI scores 72 out of 100 (grade C) on the AI assistants criteria. It meets 6 of 8 criteria: open source, no ads or data sales, no training on your data, runs locally, limited chat retention and no account needed.
Brave Leo Grade C
AI assistant built into the Brave browser for chat, page summaries and writing help. Works without an account and can also use local or self-chosen models.
Based in the United States (Five Eyes). Brave Leo scores 61 out of 100 (grade C) on the AI assistants criteria. It meets 3 of 8 criteria: no training on your data, limited chat retention and no account needed.
Duck.ai Grade C
AI chat from DuckDuckGo that relays prompts to third-party models without identifying the user, with no account needed and recent chats saved only on the device.
Based in the United States (Five Eyes). Duck.ai scores 60 out of 100 (grade C) on the AI assistants criteria. It meets 5 of 12 criteria: no trackers or telemetry, tells users about requests, no training on your data, limited chat retention and no account needed.
Kagi Assistant Grade D
AI assistant from Kagi that combines several third-party language models with Kagi Search results, available with a Kagi account.
Based in the United States (Five Eyes). Kagi Assistant scores 54 out of 100 (grade D) on the AI assistants criteria. It meets 5 of 12 criteria: no trackers or telemetry, no ads or data sales, security headers, no training on your data and limited chat retention.
Venice Grade D
AI chat and image generation service that proxies prompts to open and third-party models, keeping conversation history only in the browser.
Based in the United States (Five Eyes). Venice scores 42 out of 100 (grade D) on the AI assistants criteria. It meets 4 of 12 criteria: no ads or data sales, TLS configuration, no training on your data and limited chat retention.
Mistral Vibe Grade F
AI assistant from French company Mistral AI, formerly called Le Chat, for chat, research, documents and coding tasks. Available on the web and in mobile apps.
Based in France (Nine Eyes). Mistral Vibe scores 33 out of 100 (grade F) on the AI assistants criteria. It meets 2 of 12 criteria: no ads or data sales and TLS configuration.