Overview
What Brime is, who it's for, and the 30-second mental model.
Brime is a single API for searching the web, extracting clean content from URLs, and running multi-step research with citations. It's built for developers who don't want to glue together a SERP, a scraper, a chunker, a reranker, and an LLM.
The 30-second mental model
your code
│
▼
brime gateway ──► /v1/search (SERP + ranking + optional LLM answer)
──► /v1/extract (URLs → clean markdown)
──► /v1/research (iterative agent + citations)One key. One credit balance. Three endpoints.
When to use which
magnifying-glass
Search
You have a question that needs current web data. Pick instant for snippets, basic for an answer, advanced for higher relevance.
file-lines
Extract
You already know the URLs. PDF, DOCX, HTML, SPA — all return clean markdown. Auto Chrome escalation when needed.
brain
Research
You need synthesis across multiple sources with citations. basic for ~10-30 s sync, deep for multi-round async (1–10 min).
Hello, Brime
Three lines to your first call:
from brime import Brime
client = Brime(api_key="sk-brime-...")
print(client.search(query="Tesla Cybertruck recall 2025").answer)import { Brime } from "@brime-ai/brime";
const client = new Brime({ apiKey: "sk-brime-..." });
console.log((await client.search({ query: "Tesla Cybertruck recall 2025" })).answer);curl https://api.brime.dev/v1/search \
-H "Authorization: Bearer sk-brime-..." \
-H "Content-Type: application/json" \
-d '{"query":"Tesla Cybertruck recall 2025"}'What makes Brime different
- One key for the full RAG stack — search + scrape + research in one place; no Tavily + Exa + Firecrawl + LangChain glue.
- Drop-in compatible with Tavily, Exa, and Parallel SDKs — change
base_url, keep your code. - Modern transport — Server-Sent Events for streaming research with
Last-Event-IDresume. - Edge-native — runs at the network edge; low latency from anywhere.
- Honest pricing — credit-based, transparent multipliers, free 1,000 credits/month.