Amm1rr/WebAI-to-API

✨ What's New

• Provider-centric architecture and unified backend routing.
• Improved browser runtime lifecycle, recovery, and stability.
• Improved authentication workflows and browser runtime usability.
• Better observability, diagnostics, and operational tooling.
• Expanded documentation and test coverage.

📋 Licensing Change

• Starting with v0.5.0, WebAI-to-API is licensed under Aapache 2.0.

📋 What's Changed

• Add Proxy Option by @2077DevWave in https://github.com/Amm1rr/WebAI-to-API/pull/43
• Add endpoint by @2077DevWave in https://github.com/Amm1rr/WebAI-to-API/pull/44
• chore: remove old Gemini models and keep Gemini 3.0 set by @Phong940253 in https://github.com/Amm1rr/WebAI-to-API/pull/64
• Fix/llxprt tool calls and sse streaming by @marmarjohnson in https://github.com/Amm1rr/WebAI-to-API/pull/66
• feat: SSE streaming support + flexible model routing for CCR integration by @gyy0592 in https://github.com/Amm1rr/WebAI-to-API/pull/67
• feat: upgrade to gemini-webapi 2.0.0 with Gemini 3 model support & Expose Gemini Gems via REST endpoints by @hiyukoim in https://github.com/Amm1rr/WebAI-to-API/pull/70
• Refactor: Provider Architecture Stabilization and Operational Safety by @Amm1rr in https://github.com/Amm1rr/WebAI-to-API/pull/74
• refactor: simplify run.py and fix shutdown sequences by @Amm1rr in https://github.com/Amm1rr/WebAI-to-API/pull/75
• + 15 more

✨ New Contributors

• @2077DevWave made their first contribution in https://github.com/Amm1rr/WebAI-to-API/pull/43
• @Phong940253 made their first contribution in https://github.com/Amm1rr/WebAI-to-API/pull/64
• @marmarjohnson made their first contribution in https://github.com/Amm1rr/WebAI-to-API/pull/66
• @gyy0592 made their first contribution in https://github.com/Amm1rr/WebAI-to-API/pull/67
• @hiyukoim made their first contribution in https://github.com/Amm1rr/WebAI-to-API/pull/70
• @Amm1rr made their first contribution in https://github.com/Amm1rr/WebAI-to-API/pull/74
• @muhammetozmen made their first contribution in https://github.com/Amm1rr/WebAI-to-API/pull/87
• Full Changelog: https://github.com/Amm1rr/WebAI-to-API/compare/v0.4.0...v0.5.0

📦 WebAI to API v0.4.0 - Release Notes

• Displayed a user message explaining how to use the `gpt4free` server.
• Resolved execution issue on Windows 11.
• Improved error handling with appropriate user-facing messages.
• Updated internal libraries and dependencies.
• ---
• Full Changelog: https://github.com/Amm1rr/WebAI-to-API/compare/v0.3.0...v0.4.0

📦 WebAI to API v0.3.0 - Release Notes

• Improved server startup information display, including available services and API endpoints.
• Added a new method using the [gpt4free v0.5.5.5](https://github.com/gpt4free/g4f.dev) library, which also functions as a fallback.
• Introduced support for switching between models using keyboard shortcuts (keys `1` and `2`) in the terminal.
• WebAI-to-API now uses your browser and cookies only for Gemini, resulting in faster performance.
• `gpt4free` integration provides access to multiple providers (ChatGPT, Gemini, Claude, DeepSeek, etc.), ensuring continuous availability of various models.
• Updated internal libraries.
• Upgraded to [Gemini API v1.14.0](https://github.com/HanaokaYuzu/Gemini-API).
• Ensured compatibility with Windows (tested on Windows 11).
• + 1 more

📦 WebAI to API v0.2.3 - Release Notes

• Docker Support Added: Introduced Docker configuration for easier deployment and environment setup.
• Poetry Installation Fix: Resolved a bug related to project installation when using Poetry.
• Full Changelog: https://github.com/Amm1rr/WebAI-to-API/compare/v0.2.2...v0.2.3

📦 WebAI to API v0.2.2 - Release Notes

• AI Model Removal: Claude and Deepseek API support removed.
• Enhanced Cookie Handling: Improved cookie management for multiple cookies.
• New API Endpoints: Introduced `/translate`, `/gemini-chat`, and `/translate` endpoints.
• Maintenance: Cleaned up project, removed unused files, and updated `.gitignore`.
• Full Changelog: https://github.com/Amm1rr/WebAI-to-API/compare/v0.2.1...v0.2.2

📦 What's Fresh in WebAI to API v0.1.5

• + Project Makeover: Enjoy a revamped project structure that's now modular for easier navigation.
• + Updated Examples: Dive into refreshed examples that showcase the latest features and capabilities.
• + Clarified Documentation: Find clearer instructions and documents to guide you through every step.

📋 What's Changed

• Enhancement: add retry and waiting in the conversation creation by @linkedlist771 in https://github.com/Amm1rr/WebAI-to-API/pull/31

✨ New Contributors

• @linkedlist771 made their first contribution in https://github.com/Amm1rr/WebAI-to-API/pull/31
• Full Changelog: https://github.com/Amm1rr/WebAI-to-API/compare/v0.1.4...v0.1.5

✨ What's New in WebAI to API v0.1.4

• + Updated Google Gemini
• + Updated Claude 3
• + Removed ChatGPT
• + Added Streaming Response to Claude and Gemini
• + Fixed documents and instructions

✨ What's new in WebAI to API v0.1.3

• This release represents the first stable version with support for the recent enhancements in Claude and Bard services.

✨ New Features

• ChatGPT Code Review Integration: I have added a new `/CodeReview` endpoint to enable [ChatGPT Code Review](https://github.com/Amm1rr/ChatGPT-Code-Review) functionality. This allows users to get AI-assisted code reviews from the local server.
• ShellGPT Endpoint: A new `/ShellGPT` endpoint has been added to support [ShellGPT](https://github.com/TheR1D/shell_gpt).
• Improved Stability: I have made various fixes and improvements to increase the overall stability and robustness of the system.
• Performance Optimization for Bard: Improve connection performance with request session pool. This avoids creating a new connection for every request, improving
• overall performance.

📋 Changes

• Unified API endpoints allow seamlessly querying ChatGPT, Google Bard, and Claude from one integration.
• Leverages the free web/browser versions of each AI service by managing cookies and sessions. Avoid costs of paid API access.
• Simple API for text completion, question answering, and conversational queries. Integrate AI capabilities easily.
• Lightweight Python server using FastAPI. Easily self-host locally or on your own infrastructure.
• Full support for streaming responses.