Cerul

The video search layer for AI agents.

Teach your AI agents to see. Search video by meaning — across speech, visuals, and on-screen text.

Docs · API Reference · Pricing ·

demo.mp4

Why Cerul

Web pages are easy for AI agents to search. Video is not.

Most video search today is limited to transcripts — what was said. Cerul goes further by indexing what is shown: slides, charts, product demos, code walkthroughs, whiteboards, and other visual evidence.

Note

Cerul is in active development. The API is live at cerul.ai — sign up to get a free API key.

Quickstart

Use with your AI agent

Copy this and send it to your agent (Claude Code, Codex, Cursor, etc.):

Install the Cerul video search skill: mkdir -p ~/.claude/skills/cerul && curl -fsSL https://raw.githubusercontent.com/cerul-ai/cerul/main/skills/cerul/SKILL.md -o ~/.claude/skills/cerul/SKILL.md && cat ~/.claude/skills/cerul/SKILL.md

Your agent will install the CLI, set up credentials, and start searching videos as a tool.

Claude Code using Cerul skill to search videos

Agent synthesizing video evidence into an answer

Or install via the skills CLI:

npx skills add cerul-ai/cerul

Install for specific agents

# Claude Code
mkdir -p ~/.claude/skills/cerul && curl -fsSL https://raw.githubusercontent.com/cerul-ai/cerul/main/skills/cerul/SKILL.md -o ~/.claude/skills/cerul/SKILL.md

# Windsurf
mkdir -p ~/.codeium/windsurf/skills/cerul && curl -fsSL https://raw.githubusercontent.com/cerul-ai/cerul/main/skills/cerul/SKILL.md -o ~/.codeium/windsurf/skills/cerul/SKILL.md

# OpenCode
mkdir -p ~/.config/opencode/skills/cerul && curl -fsSL https://raw.githubusercontent.com/cerul-ai/cerul/main/skills/cerul/SKILL.md -o ~/.config/opencode/skills/cerul/SKILL.md

# Cline / Cursor
mkdir -p .claude/skills/cerul && curl -fsSL https://raw.githubusercontent.com/cerul-ai/cerul/main/skills/cerul/SKILL.md -o .claude/skills/cerul/SKILL.md

CLI

curl -fsSL https://cli.cerul.ai/install.sh | bash
cerul search "Sam Altman on AI video generation tools"

Inline video frame previews are supported in iTerm2, WezTerm, and Kitty. Enable with cerul config and toggle Images on. Other terminals show text-only results.

MCP (Model Context Protocol)

Connect any MCP-compatible client to the hosted endpoint (replace with your API key):

# Claude Code
claude mcp add --transport streamable-http "https://api.cerul.ai/mcp?apiKey=YOUR_API_KEY" cerul

# Codex
codex mcp add --url "https://api.cerul.ai/mcp?apiKey=YOUR_API_KEY" cerul

Also works with Claude Desktop, Cursor, Windsurf, and other MCP clients.

SDK & API

Python

pip install cerul

from cerul import Cerul

client = Cerul(api_key="YOUR_API_KEY")
results = client.search(query="Sam Altman on AGI timeline", max_results=5)

for r in results:
    print(r.title, r.url)

JavaScript

npm install cerul

import { cerul } from "cerul";

const client = cerul({ apiKey: "YOUR_API_KEY" });
const result = await client.search({ query: "Sam Altman on AGI timeline", max_results: 5 });

for (const r of result.results) {
  console.log(r.title, r.url);
}

cURL

curl "https://api.cerul.ai/v1/search" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "Sam Altman on AGI timeline", "max_results": 5}'

Full API spec: openapi.yaml | cerul.ai/docs

License

Licensed under Apache 2.0.

_{Built by @JessyTsui}

cerul

Description

README