[Integrations](https://spiral.ai/integrations) › [opencode](https://spiral.ai/integrations/opencode) › [Firecrawl web research](https://spiral.ai/integrations/opencode/firecrawl-web-research)

# How to Use Firecrawl web research in opencode

Firecrawl web research returns structured JSON gathered from websites it searches and navigates. Use it to assemble research across pages when finding the sources is part of the job.

|  |  |
| --- | --- |
| Model | Firecrawl web research |
| Available through | Firecrawl |
| Category | Web |
| Price | $0.001746 per unit of work the agent does, up to the run's limit ($4.36475 by default). |
| Returns | json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Firecrawl web research in opencode

- Use Firecrawl web research to find the official Next.js migration instructions relevant to the version in package.json. Create a checklist with affected APIs and source URLs.
- Use Firecrawl web research to gather React's official guidance on useEffect cleanup. Compare it with this project's effects and propose changes supported by the documentation.
- Use Firecrawl web research to collect the documented webhook retry behavior of Stripe and GitHub. Save a comparison with source URLs in docs/webhook-retries.md.
- Use Firecrawl web research to find PostgreSQL's documented JSON indexing options. Compare them with our database migrations and suggest indexes for the queries in this repository.
- Use Firecrawl web research to collect 20 recent arXiv papers about retrieval-augmented generation, including titles, abstracts, dates, and URLs. Save the returned records as a JSON fixture for our literature browser.

## How to set up Firecrawl web research in opencode

Pick the way that suits you. Each takes about two minutes.

### Use Spiral's MCP server

1. **Add the server**: Leave out `--global` to add Spiral to the current project only. No header and no client ID or secret: opencode registers itself with Spiral.

```bash
opencode mcp add spiral --global --url https://api.spiral.ai/mcp
```

2. **Sign in**: Your browser opens. Sign in to Spiral, pick the organization and click Authorize, then return to the terminal. No API key. Inside opencode, `/mcps` starts the same sign-in.

```bash
opencode mcp auth spiral
```

3. **Check it**: Spiral shows as connected.

```bash
opencode mcp list
```

### Use the Spiral skill

1. **Send opencode this message**: Your agent reads Spiral's skill and sets itself up.

```
set up https://spiral.ai/SKILL.md
```

2. **Sign in or paste a key**: Approve Spiral in the browser, or give the agent a key from the console.

### Use the Spiral CLI

1. **Install and sign in**: Installs the Spiral CLI, signs you in and adds its skill to your coding agents.

```bash
npx -y spiral-cli@latest init
```

### Call the Spiral API

1. **Create an API key**: In the console, under API keys.

2. **Run Firecrawl web research**: One request; the result comes back by polling or a webhook.

```bash
curl https://api.spiral.ai/v1/runs \
  -H "Authorization: Bearer $SPIRAL_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{
    "provider": "firecrawl",
    "model": "firecrawl/agent",
    "input": {
      "prompt": "<prompt>"
    }
  }'
```

Then paste one of the prompts above: opencode finds Firecrawl web research on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/opencode

## Firecrawl web research tools and prices

- [Firecrawl Agent · Web Research](https://spiral.ai/tools/firecrawl/firecrawl/agent) (Firecrawl): $0.001746 per unit of work the agent does, up to the run's limit ($4.36475 by default).

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| mode | extract, chat | extract | `extract` returns the complete structured result in `data` every turn. `chat` lets a follow-up that asks for no new data get a short reply in `message` instead of a re-run. `exchange.requireApproval` needs `chat` on the same request. Omitted on a follow-up turn keeps the thread's mode. |
| safe |  | — | When `true`, every web search the agent makes during the run filters explicit content, the same filter as the search endpoint's `safe` option. Pages the agent scrapes or opens in a browser are not filtered. Omit to keep the default behavior, which does not apply the filter. Omitted on a follow-up turn keeps the thread's setting; `false` turns it off. |
| urls |  | — | Optional list of URLs to constrain the agent to |
| model | spark-2, spark-1-mini, spark-1-pro | spark-2 | The model to use for the agent task. spark-2 is the default and the model every run executes on. The Spark 1 model names remain accepted for backwards compatibility but are deprecated and route to spark-2. |
| effort | low, medium, high | — | Reasoning budget for the agent task. Every run executes on spark-2, so effort can be sent with or without model. |
| schema |  | — | Optional JSON schema to structure the extracted data |

## What can you do with Firecrawl web research in opencode?

Build a migration checklist from the documentation that applies to your project's dependencies. Firecrawl can discover relevant pages and return structured findings for the coding agent to compare with project files.

### Compare documented product features

Gather feature descriptions from Notion, Asana, and Trello's public product pages, then build a comparison with supporting URLs.

### Prepare a software migration

Collect migration instructions and documented breaking changes from Next.js documentation before planning updates to an application.

### Build a focused research bibliography

Collect titles, abstracts, publication dates, and links for a bounded set of arXiv papers on retrieval-augmented generation.

### Check a public product catalog

Extract product names, specifications, and listed availability from public manufacturer pages for a purchasing shortlist.

### Maintain a company directory

Research company names, official websites, product descriptions, and industries for a defined Y Combinator batch.

### Follow product announcements

Collect dated announcements from public company blogs and compare each scheduled collection with earlier results.

### Compare visitor information

Gather opening hours, locations, and accessibility information from official Chicago museum websites for a trip-planning brief.

## Overview

- [Firecrawl](https://spiral.ai/providers/firecrawl)
- [All opencode guides](https://spiral.ai/integrations/opencode)

## Firecrawl web research in other agents

- [Amp](https://spiral.ai/integrations/amp/firecrawl-web-research)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/firecrawl-web-research)
- [Claude](https://spiral.ai/integrations/claude-desktop/firecrawl-web-research)
- [Claude Code](https://spiral.ai/integrations/claude-code/firecrawl-web-research)
- [Cline](https://spiral.ai/integrations/cline/firecrawl-web-research)
- [Codex](https://spiral.ai/integrations/codex/firecrawl-web-research)
- [Cursor](https://spiral.ai/integrations/cursor/firecrawl-web-research)
- [Devin](https://spiral.ai/integrations/devin/firecrawl-web-research)
- [dots](https://spiral.ai/integrations/dots/firecrawl-web-research)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/firecrawl-web-research)
- [Gemini](https://spiral.ai/integrations/gemini/firecrawl-web-research)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/firecrawl-web-research)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/firecrawl-web-research)
- [goose](https://spiral.ai/integrations/goose/firecrawl-web-research)
- [Grok](https://spiral.ai/integrations/grok/firecrawl-web-research)
- [Grok Build](https://spiral.ai/integrations/grok-build/firecrawl-web-research)
- [Hermes](https://spiral.ai/integrations/hermes/firecrawl-web-research)
- [Instinct](https://spiral.ai/integrations/instinct/firecrawl-web-research)
- [Manus](https://spiral.ai/integrations/manus/firecrawl-web-research)
- [Muse](https://spiral.ai/integrations/muse/firecrawl-web-research)
- [Muse Code](https://spiral.ai/integrations/muse-code/firecrawl-web-research)
- [n8n](https://spiral.ai/integrations/n8n/firecrawl-web-research)
- [Notion](https://spiral.ai/integrations/notion/firecrawl-web-research)
- [OpenClaw](https://spiral.ai/integrations/openclaw/firecrawl-web-research)
- [Perplexity](https://spiral.ai/integrations/perplexity/firecrawl-web-research)
- [Raycast](https://spiral.ai/integrations/raycast/firecrawl-web-research)
- [VS Code](https://spiral.ai/integrations/vscode/firecrawl-web-research)
- [Windsurf](https://spiral.ai/integrations/windsurf/firecrawl-web-research)
- [Zapier](https://spiral.ai/integrations/zapier/firecrawl-web-research)

## More opencode guides

- [Firecrawl web page content](https://spiral.ai/integrations/opencode/firecrawl-web-page-content)
- [Firecrawl web search](https://spiral.ai/integrations/opencode/firecrawl-web-search)
- [Firecrawl developer search](https://spiral.ai/integrations/opencode/firecrawl-developer-search)
- [Firecrawl website URLs](https://spiral.ai/integrations/opencode/firecrawl-website-urls)
- [Firecrawl batch web scraping](https://spiral.ai/integrations/opencode/firecrawl-batch-web-scraping)
- [Firecrawl website content](https://spiral.ai/integrations/opencode/firecrawl-website-content)

## Questions about Firecrawl web research in opencode

**Can opencode use Firecrawl web research?**

Yes. Add Spiral to opencode as an MCP server, and opencode can find and run Firecrawl web research like any built-in tool.

**How much does Firecrawl web research cost in opencode?**

$0.001746 per unit of work the agent does, up to the run's limit ($4.36475 by default). You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your opencode plan is separate.

**Do I need a Firecrawl account to use Firecrawl web research in opencode?**

No. Spiral runs Firecrawl web research on its own Firecrawl account. You sign in to Spiral once and never paste a key.

**How long does it take to set up Firecrawl web research in opencode?**

About two minutes: add the server, then sign in, then check it.

**Does opencode wait while Firecrawl web research runs?**

Firecrawl web research runs in the background. opencode starts the run, checks on it and gives you the json when it's ready.

**Where does Firecrawl get its research data?**

Agent searches and navigates across websites to gather the requested information. Its documented scope includes discovering sources across multiple domains and pages.

**Can Firecrawl read listings spread across multiple pages?**

Firecrawl describes Agent as handling pagination, infinite scroll, and dynamic JavaScript content. That capability does not establish that a particular extraction contains every listing.

**How many records can one research request return?**

Firecrawl documents roughly 150 to 200 structured rows per Agent run. Treat that as an approximate output size when planning a collection.

**How fresh and accurate is the information?**

Firecrawl says it fetches live pages at query time by default. Its terms still disclaim responsibility for third-party content's timeliness and accuracy, so a newly retrieved page can contain outdated information.

**Which regional version of a website does Firecrawl see?**

Firecrawl Agent browses from the US by default. Specify the intended market in research requests where regional availability matters.

