Make every bit of information verifiable to everyone.
VerifAIble is an AI-powered content creation platform where every AI response is backed by verifiable evidence. Unlike traditional AI chatbots that generate plausible but unverifiable text, VerifAIble's Agent system actively searches the web, collects evidence, and creates citations that link directly to highlighted source material — making AI-generated content auditable and trustworthy.
🌐 Try it at ai.verifaible.space
VerifAIble's AI doesn't just generate text — it takes action. The built-in Agent autonomously searches the web, navigates pages, extracts data, and creates citations through a complete tool loop (reasoning → tool calls → result feedback → continued reasoning). It supports progressive tool discovery and sub-agent orchestration for complex multi-step tasks.
Switch between 12+ LLMs from 7 providers with one click: OpenAI (GPT-5.1/5.2), Google Gemini (3-Pro/3-Flash), DeepSeek (Chat/R1), Anthropic (Claude Sonnet 4.6), MiniMax (M2.5), Zhipu (GLM-5), and Moonshot (Kimi-K2.5). Each model streams responses in real-time via SSE.
Every claim can be traced back to its source. Click any citation to see the original content highlighted on the source page. Supports 7 evidence types:
| Type | How It Works |
|---|---|
| Text | Highlights the exact sentence on the source page |
| Table | Crosshair positioning on the exact row and column |
| Opens built-in PDF viewer, jumps to the right page and passage | |
| Video | Jumps to the exact timestamp in a YouTube video |
| Dynamic Page | Replays browser actions (clicks, date pickers, JS) then highlights |
| Document | Links to uploaded documents with anchor navigation |
| Image | Links to source images with metadata |
No need to switch between apps. VerifAIble includes a built-in Chromium browser with split-view — chat with AI on one side, browse and verify sources on the other. Organize your research with workspaces and tabs.
Available on macOS, Windows, Android, and the Web. Desktop apps built with Electron, mobile with Capacitor, all sharing the same Vue 3 frontend.
Full UI support for English, 简体中文, and 日本語.
Coming soon
Note: This app is not code-signed. On first launch, you may need to allow it in System Preferences > Security & Privacy.
| Project | Description |
|---|---|
| verifaible-bench | Benchmark for evaluating LLM agents on verifiable evidence collection (6 models, 21 test cases) |
| verifaible-model | SFT fine-tuning for evidence collection capabilities |
MIT
© 2026 VerifAIble