Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the MultiOn Web API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fmultion.ai%2Fmultion" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fmultion.ai%2Fmultion" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with MultiOn Web API.
Start a browsing session that keeps page state across steps
Issue a browse command that navigates and acts on a live web page from a natural-language instruction
Retrieve structured data from the page the session is on
Capture a screenshot of the current session
List active sessions or close one when the task is done
GET STARTED
Patterns agents use MultiOn Web API for, with concrete tasks.
★ AI Agent Web Automation
Give an AI agent hands on the live web. The agent discovers the browse and session operations through Jentic, stores the MultiOn key once, and sends natural-language instructions that navigate and act on real pages, without scripting selectors. Session state carries the task across several steps.
Start a session, instruct the agent to complete a multi-step task on a target site, and return the outcome
Web Data Extraction
Pull structured data from pages that need interaction to reach. After a browse command navigates to the right view, the retrieve operation returns structured content from the current page, so an agent can gather data behind logins, forms, or dynamic navigation.
Browse to a target page, then retrieve the structured data the workflow needs from the current view
Session-Based Browsing Workflows
Run long tasks that span several page interactions. A session preserves browsing state, so an agent can issue successive browse commands, check progress, and close the session when finished, keeping each task isolated.
Open a session, run a sequence of browse steps toward a goal, and close the session once the task completes
Visual Verification
Confirm what the agent is seeing. The screenshot operation captures the current state of a session, so a workflow can log a visual record or verify a page rendered as expected before acting.
Capture a screenshot of the active session and attach it to the task log for verification
7 endpoints — the multion web api drives an autonomous browser that acts on live websites from natural-language instructions.
METHOD
PATH
DESCRIPTION
/browse
Navigate and act on a web page
/retrieve
Retrieve structured data from the current page
/session
Start a browsing session
/sessions
List active sessions
/screenshot/{session_id}
Capture a session screenshot
/browse
Navigate and act on a web page
/retrieve
Retrieve structured data from the current page
/session
Start a browsing session
/sessions
List active sessions
/screenshot/{session_id}
Capture a session screenshot
What agents get from Jentic-routed access to this vendor.
Setup
Wiring the MultiOn Web API by hand means learning its header-key auth and managing browsing sessions and their state yourself. Through Jentic you install once, import the API from the API Directory, store the key, and your agent calls the browse or retrieve operation it needs.
Permission scoping
You decide which MultiOn operations your agent may call. A rule can grant browsing and retrieval while withholding session deletion, so the agent works pages without tearing down state, and the rule bounds operations rather than a single session.
Credential isolation
Your MultiOn API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
Intent-based discovery
Agents search Jentic by intent such as 'browse a website and complete a task', and Jentic returns the matching MultiOn operation with its input schema so the agent calls the right endpoint without reading the reference docs.
Alternatives and complements available in the Jentic catalogue.
Specific to using MultiOn Web API through Jentic.
What authentication does the MultiOn Web API use?
The MultiOn Web API expects an API key sent in a request header, as declared by its OpenAPI spec. Through Jentic the key is stored encrypted by your own Jentic One instance and injected when a call runs, so it never enters the agent's prompt or logs.
Can the MultiOn Web API act on pages, not just read them?
Yes. A browse command navigates and interacts with a live page from a natural-language instruction, so the agent can click, type, and move through a flow, then use the retrieve operation to read structured data from the resulting page.
What are the rate limits for the MultiOn Web API?
The OpenAPI spec does not publish specific rate limits. Check the MultiOn documentation for current limits, and have your agent back off and retry when it receives a throttling response.
How do I automate a web task with the MultiOn Web API through Jentic?
Search Jentic for an intent like 'browse a website and complete a task', and it returns the matching MultiOn operation with its input schema. Import the connector once, store your key, and your agent drives the browser without hand-written HTTP code. To run it on your own infrastructure, install Jentic One from its GitHub repo.
Is there a MultiOn MCP server?
You don't need an MCP server to give your agent the MultiOn Web API. Jentic connects it directly from the Jentic API Directory: import the API, store your key once, and your agent calls the browse, retrieve, and session operations. Operations are discovered on demand, so no extra tool definitions are loaded into the agent's context.
Can I limit what my agent is allowed to do with the MultiOn Web API?
Yes. You choose the operations the agent may call, so a rule can allow browsing and data retrieval while withholding session deletion unless you add it. A rule bounds which operations run rather than which individual session is affected, and every call is logged.
Know of an official OpenAPI document? Contribute it →
For Agents
Drive an autonomous browser: start a session, browse and act on live web pages from natural-language instructions, retrieve page data, capture screenshots, and manage sessions.
Use for: I want to browse a website and complete a task, Start a new browsing session, Retrieve data from the current page, Take a screenshot of a session
Not supported: Does not host models or provide standalone LLM inference. Use it to drive a browsing session, act on live pages, retrieve page data, and capture screenshots only.
The MultiOn Web API drives an autonomous browser that acts on live websites from natural-language instructions. It starts a browsing session, issues a browse command that navigates and interacts with a page, retrieves structured data from the current page, and captures a screenshot. Each session keeps browsing state, and the API can list active sessions or close one when a task is finished.
This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.
Base layer of spec validity and structural soundness.
Aggregated quality score from linter diagnostics, weighted by severity.
Percentage of `$ref` references that resolve successfully.
Checks whether the API description parses successfully and conforms to its declared specification (e.g., OpenAPI).
Structural correctness score based on schema issues using logarithmic dampening.
Clarity, completeness, and ingestion readiness for developers and tooling.
How richly the API is illustrated with examples.
Percentage of examples that conform to their schemas.
Percentage of operations with complete response definitions (success, client error, server error).
Health of API ingestion, bundling, and resolution within Jentic pipelines.
Semantic breadth, depth, and agent comprehension for AI systems.
Coverage of descriptions across API elements.
Coverage of RFC 9457 Problem Details for error responses.
Coverage, uniqueness, and casing consistency of operationIds for AI inference.
Coverage of summaries across operations/tags/info.
Functional utility, complexity comfort, and AI orchestration readiness.
Agent comfort level based on API operational and structural complexity.
Trust, risk posture, and security compliance.
Average quality of security schemes based on authentication method strength (weakest link for OAuth2).
Findability, semantic richness, and reasoning readiness.
Clarity and depth of descriptions across API elements.
Score it yourself
Every API in the directory is allowlisted, so you can re-score it with no key required.
npx @jentic/api-scorecard-cli score <openapi-url>