Flux Search
Flux is a search API designed for AI agents, providing live web search results with reliable byline citations so agents can ground their responses in verifiable information. The platform offers both a chat interface for testing and API access for integration, with usage-based pricing starting at $1 free credit. Flux also includes a Lineage feature that extracts only the relevant passages from a page or document, complete with names and dates filled in.
- Artificial Intelligence
- AI Agents
- Developer Tools
- Software Only
Funding
Founders
Product
Problem
AI agents often generate responses based on outdated or unverified information, leading to hallucinations and unreliable outputs. Standard search APIs return raw links without context, forcing agents to parse entire pages and struggle to identify credible sources or distinguish bias in the information they retrieve.
Solution
Flux provides a grounded search API specifically built for AI agents, delivering live web results with byline citations that help agents assess source credibility and recency. The platform includes a Lineage feature that extracts only the passages from a page or document that directly answer a query, with names and dates filled in for immediate context. Developers can integrate Flux through a simple API, with SDKs available for OpenAI, Vercel, and LlamaIndex, or test it through a built-in chat interface. The service supports tracking companies, reviewing case law, following markets, monitoring research, and staying current on news, all with ranked results and optional passage-level answers.
Target Audience
Primary customers are developers and teams building AI agents and applications that require real-time, verifiable web search capabilities, including those working with LLM frameworks like OpenAI, Vercel, and LlamaIndex.
Features
- Live web search API returning ranked results with optional passage-level answers
- Lineage feature that extracts only the passages answering a query, with names and dates filled in
- Byline citations that help agents identify source bias and weigh new information
- SDKs and integrations for OpenAI, Vercel, and LlamaIndex
- Chat interface for testing queries before API integration
- Usage-based pricing with prepaid credits, including $1 free credit on signup and $5 free credit with a card on file
- Cached input pricing at $0.50 per 1M tokens and output at $2.76 per 1M tokens for chat
- Search pricing at $4.00 per 1,000 searches