Rust Agent Quickstart

Canonical Firecrawl Rust quickstart for external agents using search, scrape, and interact.

Firecrawl Rust Agent Quickstart

This file is the canonical quickstart for external agents integrating Firecrawl via the Rust SDK. It is generated from SDK source and the OpenAPI spec.

Install

Add to your Cargo.toml:

[dependencies]
firecrawl = "2"
tokio = { version = "1", features = ["full"] }

Authenticate

use firecrawl::Client;

let client = Client::new("fc-YOUR_API_KEY")?;

For self-hosted instances:

let client = Client::new_selfhosted("http://localhost:3000", Some("fc-YOUR_API_KEY"))?;

An empty or omitted API key is allowed — scrape, search, and interact fall back to a keyless free tier (rate-limited per IP).

When To Use What

• search — Use when you start with a query and need discovery. Returns web, news, and image results with optional scraping of each result.
• scrape — Use when you already have a URL and want page content (markdown, HTML, screenshots, structured JSON, etc.).
• interact — Use when the page needs clicks, form fills, or post-scrape browser actions. Runs code or a prompt against an active browser session.

Search

Why use it

Search the web with a query and get back structured results. Optionally scrape each result page inline. Useful for discovery, research, and finding relevant URLs before scraping them in detail.

Preferred SDK method

client.search(query, options).await

Example

use firecrawl::{Client, SearchOptions};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let client = Client::new("fc-YOUR_API_KEY")?;

    let response = client.search("firecrawl web scraping API", SearchOptions {
        limit: Some(5),
        ..Default::default()
    }).await?;

    if let Some(web_results) = response.data.web {
        for result in web_results {
            println!("{:?}", result);
        }
    }

    Ok(())
}

Results are grouped under .data.web, .data.news, and .data.images.

Parameters

SearchOptions fields (all Option, derives Default):

Field	Type	Description
limit	Option<u32>	Max number of results. Default 5, max 20.
sources	Option<Vec<SearchSource>>	Which result types: Web, News, Images.
categories	Option<Vec<SearchCategory>>	Narrow search: Github, Research, Pdf.
include_domains	Option<Vec<String>>	Only include results from these domains.
exclude_domains	Option<Vec<String>>	Exclude results from these domains.
tbs	Option<String>	Google time-based search filter (e.g. "qdr:d").
location	Option<String>	Geo-target location string.
country	Option<String>	Country code (e.g. "us").
ignore_invalid_urls	Option<bool>	Skip invalid URLs instead of erroring.
timeout	Option<u32>	Timeout in milliseconds.
highlights	Option<bool>	Generate query-relevant highlights. Defaults to true.
scrape_options	Option<ScrapeOptions>	Options applied when scraping each result.
integration	Option<String>	Integration identifier.
origin	Option<String>	Auto-set by SDK if None.

Convenience method

client.search_and_scrape(query, limit).await

Searches and scrapes all results, returning Vec<Document>.

Scrape

Why use it

Fetch a single URL and get back structured page data — markdown, HTML, screenshots, extracted JSON, and more. The workhorse endpoint for turning a known URL into usable content.

Preferred SDK method

client.scrape(url, options).await

Example

use firecrawl::{Client, ScrapeOptions, Format};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let client = Client::new("fc-YOUR_API_KEY")?;

    let doc = client.scrape("https://example.com", ScrapeOptions {
        formats: Some(vec![Format::Markdown, Format::Html]),
        only_main_content: Some(true),
        ..Default::default()
    }).await?;

    if let Some(markdown) = doc.markdown {
        println!("{}", markdown);
    }

    Ok(())
}

Pass None instead of ScrapeOptions to use all defaults.

Parameters

ScrapeOptions fields (all Option, derives Default):

Field	Type	Description
formats	Option<Vec<Format>>	Output formats: Markdown, Html, RawHtml, Links, Images, Screenshot, Summary, ChangeTracking, Json, Attributes, Branding, Product, Menu, Audio, Video. Also Question(QuestionFormat) and Highlights(HighlightsFormat).
headers	Option<HashMap<String, String>>	Custom HTTP headers.
include_tags	Option<Vec<String>>	Only include these HTML tags.
exclude_tags	Option<Vec<String>>	Exclude these HTML tags.
only_main_content	Option<bool>	Strip boilerplate, return only main content.
timeout	Option<u32>	Server-side timeout in milliseconds.
wait_for	Option<u32>	Wait time in ms for page load.
mobile	Option<bool>	Emulate mobile device.
parsers	Option<Vec<ParserConfig>>	Parser configuration (e.g. PDF).
actions	Option<Vec<Action>>	Browser actions: Wait, Click, Write, Press, Scroll, Scrape, ExecuteJavascript, Screenshot, Pdf.
location	Option<LocationConfig>	Geo-location: LocationConfig { country, languages }.
skip_tls_verification	Option<bool>	Skip TLS certificate verification.
remove_base64_images	Option<bool>	Strip base64-encoded images.
fast_mode	Option<bool>	Enable fast scraping mode.
block_ads	Option<bool>	Block ads during scraping.
proxy	Option<ProxyType>	Proxy tier: Basic, Stealth, Enhanced, Auto.
max_age	Option<u32>	Max cache age in seconds.
min_age	Option<u32>	Min cache age in seconds.
store_in_cache	Option<bool>	Store the result in cache.
lockdown	Option<bool>	Serve cached only, no outbound requests.
redact_pii	Option<bool>	Redact personally identifiable information.
audit_metadata	Option<AuditMetadata>	SIEM logging attribution.
profile	Option<ProfileConfig>	Browser profile: ProfileConfig { name, save_changes }.
integration	Option<String>	Integration identifier.
json_options	Option<JsonOptions>	JSON extraction config: schema, system_prompt, prompt, check_prompt_injection.
screenshot_options	Option<ScreenshotOptions>	Screenshot config: full_page, quality, viewport.
change_tracking_options	Option<ChangeTrackingOptions>	Change tracking config.
attribute_selectors	Option<Vec<AttributeSelector>>	Attribute extraction selectors.
origin	Option<String>	Auto-set by SDK if None.

Convenience method

client.scrape_with_schema(url, schema_json, prompt).await

Scrapes with JSON extraction using a JSON Schema value.

Interact

Why use it

Run code or a natural-language prompt against an active browser session tied to a scrape job. Use it for clicking buttons, filling forms, navigating multi-step flows, or extracting data that requires browser interaction after the initial scrape.

Preferred SDK method

client.interact(job_id, options).await

Example

use firecrawl::{Client, ScrapeExecuteOptions, ScrapeExecuteLanguage};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let client = Client::new("fc-YOUR_API_KEY")?;

    // First scrape to get a job ID
    let doc = client.scrape("https://example.com", None).await?;
    let job_id = doc.metadata.as_ref()
        .and_then(|m| m.additional.get("jobId"))
        .and_then(|v| v.as_str())
        .expect("no job ID");

    // Then interact with the browser session
    let result = client.interact(job_id, ScrapeExecuteOptions {
        code: Some("document.querySelector('button').click()".into()),
        language: Some(ScrapeExecuteLanguage::Node),
        timeout: Some(30),
        ..Default::default()
    }).await?;

    println!("{:?}", result.output);

    Ok(())
}

Parameters

ScrapeExecuteOptions fields (all Option, derives Default):

Field	Type	Description
code	Option<String>	Code to execute. At least one of code or prompt is required.
prompt	Option<String>	Natural-language prompt. At least one of code or prompt is required.
language	Option<ScrapeExecuteLanguage>	Execution language: Python, Node, Bash. Defaults to Node.
timeout	Option<u32>	Execution timeout in seconds.
origin	Option<String>	Auto-set by SDK if None.

Stopping a session

client.stop_interaction(job_id).await?;

Notes

• Async only — All methods are async and require a tokio runtime (#[tokio::main]).
• Option wrapping — All options structs use Option<T> for every field and derive Default. Use struct update syntax: ScrapeOptions { formats: Some(vec![...]), ..Default::default() }.
• No builder pattern — Options are plain structs with public fields, not builders.
• impl Into<Option<...>> — scrape() and search() accept impl Into<Option<ScrapeOptions/SearchOptions>>, so you can pass None directly or pass the struct without wrapping in Some().
• impl AsRef<str> — URL, query, and job_id parameters accept impl AsRef<str>, so &str, String, etc. all work.
• snake_case fields — Rust fields use snake_case (e.g. only_main_content). The SDK handles serde camelCase conversion for the API wire format.
• Deprecated aliases — scrape_execute() maps to interact(). stop_interactive_browser() and delete_scrape_browser() map to stop_interaction(). Always use the preferred names.

Source Of Truth

• firecrawl/apps/rust-sdk/src/v2/client.rs
• firecrawl/apps/rust-sdk/src/v2/search.rs
• firecrawl/apps/rust-sdk/src/v2/scrape.rs
• firecrawl/apps/rust-sdk/Cargo.toml
• firecrawl-docs/api-reference/v2-openapi.json