tavily-research
tavily-ai/skills
Conduct comprehensive AI-powered research with citations via Tavily CLI.
What is tavily-research?
Tavily research performs deep multi-source analysis and produces cited reports, taking 30–120 seconds. Use it when you need synthesis across sources, comparisons, market analysis, or literature reviews—not for quick fact-finding.
- Gathers and analyzes multiple web sources automatically
- Produces structured reports with explicit citations in multiple formats (numbered, MLA, APA, Chicago)
- Supports async workflow with status polling for long-running queries
- Offers model selection (mini for targeted, pro for comprehensive, auto for adaptive)
- Streams results in real-time with --stream flag
- Outputs JSON for agent integration and custom schema support
How to install tavily-research
npx skills add https://github.com/tavily-ai/skills --skill tavily-research- Tavily CLI installed (npx skills add tavily-cli if missing)
- Tavily API authentication via tvly login or TAVILY_API_KEY environment variable
How to use tavily-research
- 1.Run tvly research "your query" to start a basic research task
- 2.Choose --model mini for single-topic queries or --model pro for complex comparisons
- 3.Use --stream to watch progress in real-time, or --no-wait to get a request_id for async polling
- 4.Specify --citation-format (numbered, mla, apa, chicago) if you need a particular citation style
- 5.Save output with -o filename.json or use --json for structured agent-readable results
Use cases
- Compare competing products or frameworks (e.g., AI code assistants, EV manufacturers)
- Analyze market trends and competitive landscapes
- Conduct literature reviews on technical or business topics
- Synthesize multi-angle analysis on complex questions like 'best practices for X'
- Generate cited reports for research or decision-making
- Researchers and analysts needing multi-source synthesis
- Product managers evaluating competitive landscapes
- Engineers investigating technical trends or frameworks
- Anyone conducting in-depth analysis requiring citations
tavily-research FAQ
30–120 seconds depending on query complexity and model choice. Use --stream to see progress.
Use research for deep synthesis, comparisons, and multi-source analysis. Use tavily-search for quick fact-finding.
Yes: use --no-wait to get a request_id, then tvly research poll <request_id> to check results later.
Mini is faster (~30s) for single-topic queries; pro is comprehensive (~60–120s) for comparisons and multi-faceted analysis.
Use --json for structured output, or --output-schema with a custom JSON schema for tailored formatting.
Full instructions (SKILL.md)
Source of truth, from tavily-ai/skills.
name: tavily-research description: | Conduct comprehensive AI-powered research with citations via the Tavily CLI. Use this skill when the user wants deep research, a detailed report, a comparison, market analysis, literature review, or says "research", "investigate", "analyze in depth", "compare X vs Y", "what does the market look like for", or needs multi-source synthesis with explicit citations. Returns a structured report grounded in web sources. Takes 30-120 seconds. For quick fact-finding, use tavily-search instead. allowed-tools: Bash(tvly *)
tavily research
AI-powered deep research that gathers sources, analyzes them, and produces a cited report. Takes 30-120 seconds.
Before running
Research requires authentication. Run the requested command directly when
tvly is already authenticated; do not add a status check to every invocation.
If tvly is missing, follow the tavily-cli setup.
If an installed CLI reports an authentication error, use tvly login for
authentication only, or tvly init --skip-skills when guided verification is
also useful. Browser-based OAuth is preferred when an interactive user can
complete it. --no-browser prints the sign-in link instead of opening it, but
still waits for a localhost callback. In an unattended agent or CI environment,
leave authentication to the user or use a securely provided TAVILY_API_KEY.
Do not start a second login immediately after guided setup has completed.
When to use
- You need comprehensive, multi-source analysis
- The user wants a comparison, market report, or literature review
- Quick searches aren't enough — you need synthesis with citations
- Step 5 in the workflow: search → extract → map → crawl → research
Quick start
# Basic research (waits for completion)
tvly research "competitive landscape of AI code assistants"
# Pro model for comprehensive analysis
tvly research "electric vehicle market analysis" --model pro
# Stream results in real-time
tvly research "AI agent frameworks comparison" --stream
# Save report to file
tvly research "fintech trends 2025" --model pro -o fintech-report.json
# JSON output for agents
tvly research "quantum computing breakthroughs" --json
Options
| Option | Description |
|---|---|
--model | mini, pro, or auto (default) |
--stream | Stream results in real-time |
--no-wait | Return request_id immediately (async) |
--output-schema | Path to JSON schema for structured output |
--citation-format | numbered, mla, apa, chicago |
--poll-interval | Seconds between checks (default: 10) |
--timeout | Max wait seconds (default: 600) |
-o, --output | Save the JSON response to a file |
--json | Structured JSON output |
Model selection
| Model | Use for | Speed |
|---|---|---|
mini | Single-topic, targeted research | ~30s |
pro | Comprehensive multi-angle analysis | ~60-120s |
auto | API chooses based on complexity | Varies |
Rule of thumb: "What does X do?" → mini. "X vs Y vs Z" or "best way to..." → pro.
Async workflow
For long-running research, you can start and poll separately:
# Start without waiting
tvly research "topic" --no-wait --json # returns request_id
# Check status
tvly research status <request_id> --json
# Wait for completion
tvly research poll <request_id> --json -o result.json
Tips
- Research takes 30-120 seconds — use
--streamto see progress in real-time. - Use
--model profor complex comparisons or multi-faceted topics. - Use
--output-schemato get structured JSON output matching a custom schema. - For quick facts, use
tvly searchinstead — research is for deep synthesis. - Read from stdin:
echo "query" | tvly research - --json
See also
- tavily-search — quick web search for simple lookups
- tavily-crawl — bulk extract from a site for your own analysis
Related skills
More from tavily-ai/skills and the wider catalog.
tavily-search
Search the web with LLM-optimized results, snippets, and relevance scores via Tavily CLI.
crawl
Crawl websites and save pages as markdown files for offline access and analysis.
extract
Extract clean markdown/text content from specific URLs using Tavily's extraction API.
research
Comprehensive web research with automatic source gathering and explicit citations.
drawio-flowchart
Creates flowcharts, swim lane diagrams, and business process diagrams using Draw.io (diagrams.net). Covers standard flowchart shapes (process, decision, start/end), connectors, auto-layout, and export to PNG/SVG/PDF. Use when the user needs to visualize workflows, decision trees, or business processes.
uniapp-project
Provides per-component and per-API examples with cross-platform compatibility details for uni-app, covering built-in components, uni-ui components, and APIs (network, storage, device, UI, navigation, media). Use when the user needs official uni-app components or APIs, wants per-component examples with doc links, or needs platform compatibility checks.