anycapanycap
Capabilities

Generate

Image GenerationCreate and edit images from prompts or references.Video GenerationCreate motion outputs from text and image inputs.Music GenerationProduce music tracks through one runtime.Audio GenerationGenerate speech, dialogue, sound effects, and complete audio scenes from text, audio, or image input.

Understand

Image UnderstandingRead screenshots, diagrams, and visual references.Video AnalysisInspect recordings and extract structured details.Audio UnderstandingTranscribe and analyze voice and audio files.

Retrieve

Web SearchSearch the web from the same agent workflow.Grounded Web SearchReturn synthesized answers with live citations.Web CrawlFetch pages and convert them into clean content.

Store

DriveStore outputs, organize assets, and create public URLs.
Equip Agents
Claude CodeCursorCodexDeepSeek HarnessManus
Resources

Explore

GuidesDecision guides for building reliable agent workflows.Context EngineeringUnderstand how prompts, files, and workspace state shape agent behavior.Agent SkillsSee how reusable skills package workflows and capability usage for agents.

Evaluate

Compare AnyCapBrowse comparison pages for adjacent agent tooling, media APIs, and tradeoffs.GlossaryA shared vocabulary for agent capabilities, tools, and workflows.
Docs ↗Pricing
I'm Agent
I'm Agent
  1. Home
  2. Capabilities
  3. Web Search

Capabilities · Live retrieval

Web Search API
for AI agents

AnyCap gives AI agents a structured web search API through one CLI. Use it for source discovery and result retrieval; choose grounded web search when the workflow needs a synthesized answer with citations, or web crawl when the URL is already known. Standard web search can constrain trusted domains, filter by recency, and optionally pull source content into the next agent step.

Equip your AgentGrounded SearchWeb Crawl

Find the sources.

The agent retrieves structured results before it writes, cites, or crawls.

✻agent with anycap
❯

I'll search the web using AnyCap.

●run("anycap search --query 'Go 1.25 release notes' --include go.dev --max-results 2 --no-crawl")

✓Search complete — 2 results from go.dev.

1. go.dev/doc/go1.25 — Go 1.25 Release Notes

2. go.dev/doc/devel/release — Release History

Search checked September 9, 2026. Titles and URLs shown; results can change.

Open the release notes for details and the release history for subsequent updates.

❯

Agents do not need another disconnected tool.
They need the capability inside the workflow.

AnyCap turns capability access into agent action.

The short answer

Use AnyCap web search when the agent needs source discovery, trusted-domain retrieval, recency filters, or a result set it can inspect before synthesis.

01

Discovery: Web search wins when the answer starts with "find me the sources" rather than "tell me the answer." It is the right first step when the agent needs options.

02

Verification: It also wins when the team cares about recency, domain trust, or a constrained source list. Those are usually the cases where a generic browser search feels too loose for production work.

03

Handoff: It loses value if you stop at the search result. The useful outcome is a source bundle that another agent, editor, or human can continue from without repeating the same search work.

Checked against the current CLI contract · CLI contract verified August 24, 2026

Structured retrieval and grounded answers are separate modes.

We checked the current AnyCap search command contract. General mode returns a result set the agent can inspect, filter, and optionally crawl. Grounding mode returns a synthesized answer with source attribution. Keeping those jobs separate prevents a retrieval page from promising an answer workflow it does not perform.

Open the CLI reference

General mode

Structured results · 1 credit

Grounding mode

Synthesized answer with citations · 5 credits

Result control

1–20 results, with optional full-page crawl

Filters

Recency, date range, include domains, exclude domains

$anycap search --query "AI agent framework" --max-results 5 --time-range month

Inspectable output

General-search response excerpt · checked September 9, 2026

{
  "status": "success",
  "data": {
    "mode": "general",
    "query": "Go 1.25 release notes",
    "results": [
      {
        "title": "Go 1.25 Release Notes - The Go Programming Language",
        "url": "https://go.dev/doc/go1.25"
      },
      {
        "title": "Release History - The Go Programming Language",
        "url": "https://go.dev/doc/devel/release"
      }
    ]
  }
}

This is a representative response shape, not a guarantee that every analysis returns the same prose. The content changes with the recording and instruction; operational fields may vary by server response.


How the workflow usually runs

01 / Step 1

Start with structured search when the agent does not know the source yet. This gives you a result set instead of a speculative answer, which is usually the right first move for research, competitive scans, and release tracking.

02 / Step 2

Use grounded search when the agent needs to turn evidence into an answer with citations. Use crawl when the exact URL is already known and you just need the page content without another discovery pass.

03 / Step 3

Hand the result set back to the agent or team as source notes, citations, or a short evidence bundle. That makes web search an input to later work instead of a dead-end feature that stops at search results.


CLI usage

General web search

$anycap search --query "AI agent framework"

Fast titles and URLs only

$anycap search --query "AI agent framework" --no-crawl

Constrain domains

$anycap search --query "Go context" --include github.com --include go.dev

Recent results only

$anycap search --query "Claude Code release" --time-range month

When agents need web search

Use case 1

Find relevant pages, docs, or references before an agent writes or updates code.

Use case 2

Search the web with structured results instead of opening a browser-based search workflow.

Use case 3

Constrain searches to trusted domains when teams need more predictable retrieval.

Use case 4

Pull page content directly into the workflow when the workflow needs more than just URLs and snippets.

Use case 5

Support research, content, and operations workflows that need source lists instead of a final answer.


Related pages

Related capability

Grounded Web Search

Move up to citation-backed answers when the workflow needs synthesis instead of retrieval.

Related capability

Web Crawl

Use crawl when the URL is already known and the workflow just needs the page content.

Pricing

Pricing

See the credit model for web search, grounded answers, and crawl requests.

Agent workflow

Web search tools for Codex

See how Codex adds search, media, storage, and publishing through one terminal-native capability runtime.

Agent workflow

Web search tools for Cursor

Keep source discovery and follow-up work inside Cursor's editor and terminal workflow.


FAQ

What does AnyCap web search let agents do?

It gives agents a structured web search API through the CLI. Teams can search by query, filter by domain or date, and optionally fetch full page content without stitching together separate browser tools or search providers.

What does a search result contain?

General search returns structured source results with titles, URLs, and descriptions, with full-page crawling enabled by default. Use --no-crawl when you only need titles and URLs. For a synthesized answer with source attribution, use grounding mode instead.

How is this different from grounded web search?

Web search returns structured search results. Grounded web search synthesizes an answer with citations. The two capabilities share the same command family but solve different jobs.

When should teams use standard web search first?

When the goal is discovery, retrieval, or collecting a result set. If the agent needs to inspect candidate pages itself, standard search is the cleaner starting point.

Is this only for developers?

No. Web search is useful for researchers, content teams, operators, and support workflows that need trusted source lists, competitive scans, or fast collection of live references.

Let your agent find the sources.

Use web search when the workflow starts with discovery, result selection, and source bundles rather than a final answer.

Equip your AgentGrounded SearchWeb Crawl

Official documentation

Implement retrieval with the official reference.

Use the Docs to configure web retrieval in your agent workflow and review the security guidance before production use.

Web capabilitySecurity reference

In practice

Read the workflow before you build it.

These implementation guides connect the capability decision to a concrete agent workflow.

  • Give AI agents web-search capability →Understand where live retrieval belongs in a reliable agent workflow.
  • Web-search APIs for AI agents →Compare the retrieval decisions that matter before an agent synthesizes or acts on sources.

Capabilities

  • Overview
  • Image Generation
  • Video Generation
  • Music Generation
  • Image Understanding
  • Video Analysis
  • Audio Understanding
  • Web Search
  • Grounded Web Search
  • Web Crawl
  • Drive

Equip Agents

  • Overview
  • Start here
  • Claude Code
  • Cursor
  • Codex
  • Manus

Resources

  • Overview
  • Context Engineering
  • Agent Skills
  • What Agents Can't Do
  • Compare agents and tools

Product

  • Product overview
  • Models
  • Install AnyCap
  • Install the Agent Skill

Documentation

  • Docs overview
  • Install AnyCap
  • MCP setup
  • CLI reference

Published on AnyCap

  • AI guides
  • Blog
  • News

Company

  • About
  • Contact
  • Privacy
  • Terms
anycap
Join the AnyCap Discord