tavily-research
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
Chinesetavily research
tavily research
AI-powered deep research that gathers sources, analyzes them, and produces a cited report. Takes 30-120 seconds.
借助AI的深度研究功能,可收集来源、分析内容并生成带引用的报告,耗时30-120秒。
Prerequisites
前提条件
Requires the Tavily CLI. See tavily-cli for install and auth setup.
Quick install:
curl -fsSL https://cli.tavily.com/install.sh | bash && tvly login需要安装Tavily CLI。请查看tavily-cli了解安装和认证设置。
快速安装:
curl -fsSL https://cli.tavily.com/install.sh | bash && tvly loginWhen to use
使用场景
- 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
- 需要全面的多来源分析
- 用户需要对比报告、市场报告或文献综述
- 快速搜索无法满足需求,需要带引用的内容整合
- 工作流的第5步workflow:搜索→提取→映射→爬取→研究
Quick start
快速开始
bash
undefinedbash
undefinedBasic research (waits for completion)
基础研究(等待完成)
tvly research "competitive landscape of AI code assistants"
tvly research "competitive landscape of AI code assistants"
Pro model for comprehensive analysis
使用Pro模型进行全面分析
tvly research "electric vehicle market analysis" --model pro
tvly research "electric vehicle market analysis" --model pro
Stream results in real-time
实时流式输出结果
tvly research "AI agent frameworks comparison" --stream
tvly research "AI agent frameworks comparison" --stream
Save report to file
将报告保存到文件
tvly research "fintech trends 2025" --model pro -o fintech-report.md
tvly research "fintech trends 2025" --model pro -o fintech-report.md
JSON output for agents
为Agent输出JSON格式结果
tvly research "quantum computing breakthroughs" --json
undefinedtvly research "quantum computing breakthroughs" --json
undefinedOptions
选项
| Option | Description |
|---|---|
| |
| Stream results in real-time |
| Return request_id immediately (async) |
| Path to JSON schema for structured output |
| |
| Seconds between checks (default: 10) |
| Max wait seconds (default: 600) |
| Save output to file |
| Structured JSON output |
| 选项 | 描述 |
|---|---|
| |
| 实时流式输出结果 |
| 立即返回request_id(异步模式) |
| 自定义结构化输出的JSON schema路径 |
| |
| 状态检查间隔秒数(默认:10) |
| 最长等待秒数(默认:600) |
| 将输出保存到文件 |
| 输出结构化JSON格式 |
Model selection
模型选择
| Model | Use for | Speed |
|---|---|---|
| Single-topic, targeted research | ~30s |
| Comprehensive multi-angle analysis | ~60-120s |
| API chooses based on complexity | Varies |
Rule of thumb: "What does X do?" → mini. "X vs Y vs Z" or "best way to..." → pro.
| 模型 | 适用场景 | 速度 |
|---|---|---|
| 单一主题、针对性研究 | ~30秒 |
| 全面的多角度分析 | ~60-120秒 |
| 由API根据内容复杂度自动选择 | 时长不定 |
经验法则:"X有什么作用?"→用mini。"X vs Y vs Z"或"最佳方法是..."→用pro。
Async workflow
异步工作流
For long-running research, you can start and poll separately:
bash
undefined对于耗时较长的研究任务,你可以分开启动和轮询:
bash
undefinedStart without waiting
启动任务无需等待
tvly research "topic" --no-wait --json # returns request_id
tvly research "topic" --no-wait --json # returns request_id
Check status
检查状态
tvly research status <request_id> --json
tvly research status <request_id> --json
Wait for completion
等待任务完成
tvly research poll <request_id> --json -o result.json
undefinedtvly research poll <request_id> --json -o result.json
undefinedTips
提示
- Research takes 30-120 seconds — use to see progress in real-time.
--stream - Use for complex comparisons or multi-faceted topics.
--model pro - Use to get structured JSON output matching a custom schema.
--output-schema - For quick facts, use instead — research is for deep synthesis.
tvly search - Read from stdin:
echo "query" | tvly research - --json
- 研究耗时30-120秒——使用可实时查看进度。
--stream - 复杂对比或多维度主题请使用。
--model pro - **使用**可获取符合自定义schema的结构化JSON输出。
--output-schema - 快速查找事实请使用——research功能适用于深度内容整合。
tvly search - 从标准输入读取:
echo "query" | tvly research - --json
See also
另请参阅
- tavily-search — quick web search for simple lookups
- tavily-crawl — bulk extract from a site for your own analysis
- tavily-search——用于简单查询的快速网页搜索
- tavily-crawl——批量提取网站内容供自行分析