Ptero.pro is a free, non-profit, open-source AI chat platform that aggregates multiple AI models and provides a chat service that can be used without registration.
Purpose The platform is aimed at users who want to use multiple AI models for conversation for free, emphasizing privacy and open source: no registration required, completely free, and the source code is publicly available for inspection.
Main capabilities The platform provides multi-model AI chat, and users can choose a model from it for conversation. According to the information, the currently configurable models include Ling 3.0 Flash (Free), Deepseek v4.1 Flash (Free), Mistral Nemo 12B, GLM 5.3 Flash, GPT-OSS 20B, Apodex 1.1 Mini (Free), and others. The platform also provides a Projects feature for categorizing and saving related chats to the cloud, and supports importing public GitHub projects. The platform provides personal API keys, which can be used to call interfaces such as chat completions, listing models, and querying usage through an OpenAI-compatible API endpoint, with support for Server-Sent Events streaming output.
How to use Users can use it directly as a guest by choosing a display name; that name and the guest identity token are stored in the browser's local storage to distinguish conversations. Users can also log in with a WordPress account or GitHub. Regular chat history is stored only locally in the browser; if cross-device sync is needed, Projects can be used to save chats, project titles, and custom files to the server and associate them with the logged-in account. Users can migrate local chat history between different devices by exporting/importing JSON backups (the media files themselves remain on the original device). For API use, a personal API key must first be created (GitHub verification is required before creation, and the key is shown only once), and the key is sent as a Bearer token to the chat endpoint. The streaming endpoint provided by the information is https://ptero.pro/wp-json/mlp/v1/chat-stream。API The documentation shows support for an OpenAI-compatible interface, with an example base URL of https://api.example.com/v1，端点包括 /chat/completions (POST), /models (GET), /usage (GET). The example model mercury-2 is provided by Inception Labs, with 50,000 tokens per hour.
Applicable scenarios and limitations Applicable scenarios include everyday AI conversations, multi-model comparison experiences, lightweight chats that require privacy protection, and integrating chat capabilities into your own applications via API. In terms of limitations: the platform does not provide paid plans; regular chats are not stored on the server, and clearing browser data or switching devices may lose local chats, so backups should be exported in advance; data within Projects is stored on the server, and project files are deleted after 90 days without being opened; API requests and web chat share the same rate limits and token quotas, and after reaching the hourly token limit, requests will be rate-limited until the top of the hour reset; AI output is generated by third-party models and may be inaccurate or incomplete, and does not constitute professional advice; the platform may suspend or disable the service, accounts, models, or guest identities at any time; the platform routes messages to third-party AI providers, whose terms and privacy practices may apply.
