Citations wired to the sentence, not the page
Inline markers map to the exact source that produced each claim. Hover one and the source card surfaces — title, domain, favicon, snippet — so verifying an answer takes a second, not a new tab.
v2 · shipped
Answers that
Answers that
show their
receipts.
Fireplexity is an open-source AI answer engine. It searches the live web, scrapes what it finds with Firecrawl, and streams back a cited answer — alongside news, images, and ticker charts.
MIT licensed
Next.js 15
Bring your own keys
how does firecrawl handle javascript-heavy pages?
Sources · 4
Ffirecrawl.dev
Ggithub.com
Ynews.ycombinator
Ddocs.firecrawl.dev
Answer
Firecrawl renders each page in a real headless browser before extraction, so client-side routes and SPA content resolve the same way they would for a visitor1. The rendered DOM is then converted to clean markdown, stripping nav and boilerplate so only the substance reaches the model24
Images
Ask next
What about rate limits?
Compare to Puppeteer
What it does
One query in.
Six kinds of answer out.
Every response is assembled from live sources, not a training snapshot — and every claim carries a link you can open.
Live web search, scraped in the same pass
Firecrawl returns ranked results and their full markdown together, so the model reads whole pages instead of guessing from search snippets.
News with dates attached
Time-sensitive questions route through a news pass that keeps outlet and publish date visible, so you can judge how stale an answer is.
Image results
Visual queries return a thumbnail grid with a full-size lightbox and the page each image came from.
Ticker detection
Mention a company and Fireplexity spots the symbol, then drops a live TradingView chart in beside the answer.
Follow-ups that go somewhere
Each answer ends with three generated next questions, drawn from what the sources actually covered.
How it works
Three hops from
question to citation.
No vector store, no nightly crawl, no index to keep warm. The pipeline runs per query, start to finish, in one request.
01
Search, three ways
The query fans out to Firecrawl's web, news, and image endpoints at once, so a single ask covers all three surfaces.
02
Scrape to markdown
Each result is rendered in a real browser, then stripped of nav, ads, and boilerplate — leaving LLM-ready markdown instead of raw HTML.
03
Synthesize and stream
Kimi K2 on Groq reads the markdown and streams an answer token by token through the AI SDK, tagging claims with the source they came from.
Why run your own
A closed answer engine
is a black box you rent.
Fireplexity is the whole thing — prompt, pipeline, and UI — in a repo you control.
Read the prompt, change the prompt
The system prompt, the citation format, the follow-up logic — all of it is a file in app/api. Nothing about how answers are shaped is hidden from you.
Your queries stay yours
Deploy it inside your own infrastructure and sensitive searches never pass through a vendor's logs or analytics.
Swap any layer
The model, the search provider, the renderer, the theme. Each sits behind its own boundary, so replacing one doesn't mean rewriting the rest.
Pay for calls, not seats
Two API keys and a Vercel deploy. No per-user pricing, no plan tier gating the features you already have the code for.
The stack
Boring parts,
on purpose.
Everything here is something you already know how to debug. No bespoke framework, no custom runtime.
Next.js 15 app router
React 19
AI SDK 5 streaming
Firecrawl v3 search + scrape
Groq inference
Kimi K2 synthesis
Tailwind v4
shadcn/ui
Streamdown
Shiki code blocks
TradingView charts
TypeScript
Self-host
Running in
about four minutes.
- Clone and installNode 18 or newer. No database, no queue, no extra services.
- Add two keysGrab them from firecrawl.dev and console.groq.com, then drop them in .env.local.
- Start itOpen localhost:3000 and ask something. Or push straight to Vercel — the repo deploys as-is.
$ git clone https://github.com/firecrawl/fireplexity.git $ cd fireplexity $ npm install # two keys, both free to start $ cp .env.example .env.local $ echo "FIRECRAWL_API_KEY=fc-..." >> .env.local $ echo "GROQ_API_KEY=gsk_..." >> .env.local $ npm run dev → ready on http://localhost:3000
Stop trusting answers
you can't check.
Fireplexity is MIT licensed and the entire pipeline is a few hundred lines. Read it, fork it, point it at your own sources.