How much does Bright Data Agent Browser cost?

Pay as you go at $8 a GB with no monthly commitment; Scale at $499 a month with 71 GB included; Enterprise custom with SSO, audit logs and a premium SLA. The pricing page shows a free tier with 1 GB a month, while Bright Data's llms.txt says the free tier excludes the Browser API and quotes from $5 a GB.

Does Bright Data Agent Browser support MCP and tool calling?

Full under the Agentic Index verification standard. Bright Data Agent Browser publishes public evidence of Model Context Protocol (MCP) support, tool calling, and API or SDK extensibility for connecting agents to external tools. Evidence basis: Sessions are driven over a CDP WebSocket endpoint at brd.superproxy.io, port 9222 for Puppeteer and Playwright and 9515 for Selenium, with zone credentials and examples in JavaScript, Python and C#. Zones are created in the control panel or. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Does Bright Data Agent Browser have SOC 2, SSO and role based access control?

Full under the Agentic Index verification standard. Bright Data Agent Browser publishes public evidence of enterprise security and identity controls, covering a compliance posture such as SOC 2 or ISO 27001, single sign on (SSO or SAML), and role based access control (RBAC). Evidence basis: SSO and audit logs come with the Enterprise plan, and each Browser API zone has its own username and password. Bright Data holds SOC 2, with the report available on contact, and offers a downloadable SOC 3 report. Its trust center also cove. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Does Bright Data Agent Browser give you tracing and observability into agent runs?

Partial under the Agentic Index verification standard. Bright Data Agent Browser publishes some public evidence of logging or observability, but narrower than full tracing and auditability across agent runs. Evidence basis: Each session returns an inspect URL to watch it live in DevTools, and the Enterprise plan adds account audit logs. Bright Data names no recorded run trace of what a session did. Bright Data, docs.brightdata.com/products/scraping-browser/con. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Back to vendors

Bright Data Agent Browser

Q: Who is Bright Data Agent Browser for?

AI agent developers, data collection teams, enterprises needing reliable web access, agent platform builders

Q: What are the best alternatives to Bright Data Agent Browser?

Based on documented capability coverage across the Agentic Index 14 point taxonomy, the closest alternatives to Bright Data Agent Browser among browser and computer-use agents are AgentQL, TinyFish, Stagehand, Notte. Each is graded on the same public evidence, and no vendor pays for placement.

Also known as: Bright Data Agent Browser, Scraping Browser, Browser API

Visit site

Browser / computer-use agentfirst party productVerified 2026-09-27

Entry priceFree tier (1 GB a month on the Browser API pricing page); pay as you go $8 a GB; Scale $499 a month with 71 GB; Enterprise customFull pricing detail

Bright Data's cloud browser runtime for AI agents (the Browser API), running headful or headless Chromium sessions with built in unblocking, CAPTCHA solving and IP rotation, driven over CDP from Puppeteer, Playwright or Selenium, with SSO and audit logs on Enterprise.

Agent Browser is Bright Data's cloud browser runtime for AI agents, which Bright Data sells as the Browser API. Developers connect Puppeteer, Playwright or Selenium to a CDP WebSocket endpoint with a zone's credentials, and Bright Data runs headful or headless Chromium sessions that handle fingerprinting, CAPTCHA solving, retries and IP rotation from its proxy network, with sessions able to appear from any country, city or ASN and unlimited sessions in parallel. Each session returns an inspect URL for live DevTools debugging, and sessions time out after 5 idle minutes or 60 minutes in all. Bright Data's free MCP server connects MCP clients to its browsing and extraction tools.

The product is browser runtime, not an agent. The customer's code or agent framework holds the task logic and the model. Enterprise adds SSO, audit logs, a premium SLA and custom packages, and Bright Data's compliance covers SOC 2 (report on request), ISO 27701 and GDPR. Bright Data's wider web data suite, including the Web Unlocker, SERP, Web Scraper and Crawl APIs, datasets and proxies, has its own Bright Data listing.

Vendor details

Canonical URL

https://brightdata.com/ai/agent-browser

Funding status

A product of Bright Data (formerly Luminati Networks), the web data infrastructure company graded on the bright-data record

Company status

first party product

Use cases & customers

Primary use cases

unblockable browser sessions for AI agentsweb scraping behind anti-bot systemsautonomous CAPTCHA solving at scalegeo targeted data collectionagent web access via MCP and AI SDK tools

Target customers

AI agent developersdata collection teamsenterprises needing reliable web accessagent platform builders

Deployment options

Cloud (Bright Data infrastructure)Headful or headless Chromium sessions

Integrations

Puppeteer, Playwright and Selenium connect to a CDP WebSocket endpoint with zone credentials, and zones are created in the control panel or by API. Browser instances integrate with an enterprise VPN and sign on, and Bright Data's free MCP server serves MCP clients.

In practice

Your research agent keeps getting blocked by CAPTCHAs on retail sites. Point it at Agent Browser over CDP, and sessions solve CAPTCHAs, rotate IPs and render the page so the agent can read and click through.

A pricing team needs to see a product page the way shoppers in Tokyo and Chicago do. Agent Browser opens sessions from the chosen city or ASN, and the team's Playwright script collects what each one sees.

Your developer is debugging why an agent stalls on a checkout form. The session's inspect URL opens live DevTools, so they can watch the run and fix the script.

Sources & related URLs

Research sources

https://brightdata.com/ai/agent-browser https://brightdata.com/blog/ai/agent-browser-vs-puppeteer-playwright https://github.com/haoruilee/awesome-agent-native-services/blob/main/services/browser-and-web-execution/bright-data-agent-browser.md

Agentic Index coverage score

4.5 / 14 capabilities · 32%

Integrations & Tool Calling	Partial
Browser instances can integrate with an enterprise VPN and sign on, which gives sessions authenticated reach into company systems. Bright Data names no connector catalog or tool route. Puppeteer, Playwright, Selenium and agent frameworks connect in through the API or MCP to drive the browser. SourceBright Data, brightdata.com/ai/agent-browserread 2026-10-06
Workflow Orchestration	Not documented
Agent Browser is browser runtime, so the customer's code or agent framework holds the task logic and drives sessions over CDP. Bright Data names no workflow model of its own. Multi step workflows and form interactions are use cases the customer scripts. Each session covers one domain, and switching domains needs a new session. SourceBright Data, docs.brightdata.com/products/scraping-browser/configuration and /introductionread 2026-10-06
Knowledge Grounding & RAG	Not documented
Pages are fetched and rendered at run time for the customer's code, including JavaScript heavy sites and dynamic content. Bright Data names no retrieval structure over the customer's knowledge for the Browser API. SourceBright Data, docs.brightdata.com/products/scraping-browser/introductionread 2026-10-06
Human Oversight & Guardrails	Not documented
CAPTCHAs, blocks and session recovery are handled automatically, and live inspection of a session is for debugging. Bright Data names no approval, takeover or guardrail mechanism. SourceBright Data, docs.brightdata.com/products/scraping-browser/configuration and /introductionread 2026-10-06
Security, Identity & Governance	Full
SSO and audit logs come with the Enterprise plan, and each Browser API zone has its own username and password. Bright Data holds SOC 2, with the report available on contact, and offers a downloadable SOC 3 report. Its trust center also covers ISO 27701, GDPR, a CSA STAR registry listing and a PwC independent assurance report on its compliance and ethics controls. SourceBright Data, brightdata.com/trustcenter, /ai/agent-browser and docs.brightdata.com/products/scraping-browser/configurationread 2026-10-06
Observability & Auditability	Partial
Each session returns an inspect URL to watch it live in DevTools, and the Enterprise plan adds account audit logs. Bright Data names no recorded run trace of what a session did. SourceBright Data, docs.brightdata.com/products/scraping-browser/configuration and brightdata.com/ai/agent-browserread 2026-10-06
Memory & State Persistence	Not documented
Managed sessions control cookies, context and tabs within a session, which times out after 5 idle minutes or 60 minutes in all. Nothing persists across sessions for an agent to read. SourceBright Data, docs.brightdata.com/products/scraping-browser/configurationread 2026-10-06
Deployment & Data Residency	Not documented
Sessions run on Bright Data's cloud, which handles 1M+ concurrent sessions by its own count. Customers can choose the country, city, state or ASN a session appears from, across 400M+ IPs in 195 countries. That choice sets proxy geolocation, not where data is processed or stored. Bright Data names no region, VPC or self hosted option. SourceBright Data, brightdata.com/ai/agent-browserread 2026-10-06
Prebuilt Agents, Templates & Packs	Not documented
The Browser API ships code examples for Puppeteer, Playwright and Selenium but no prebuilt agents or templates. The 1,300+ pre built scrapers belong to the Web Scraper API, a different Bright Data product. Usage is priced by data, from $8 per GB pay as you go down to $5 per GB, with monthly plans from $499 with 71 GB included to $1,999 with 399 GB. SourceBright Data, brightdata.com/ai/agent-browser and docs.brightdata.com/products/scraping-browser/configuration; brightdata.com/llms.txtread 2026-10-06
Triggers & Channel Coverage	Not documented
Sessions open when the customer's code connects to the CDP endpoint. Bright Data names no schedule or event trigger for the Browser API, and scheduled collection belongs to other Bright Data products. SourceBright Data, docs.brightdata.com/products/scraping-browser/introduction and /configurationread 2026-10-06
Model Flexibility & Routing	Not documented
No model is part of the Browser API. The customer's own agent and model drive the sessions through the API or MCP. SourceBright Data, docs.brightdata.com/products/scraping-browser/introduction and brightdata.com/ai/agent-browserread 2026-10-06
APIs, SDKs & MCP Extensibility	Full
Sessions are driven over a CDP WebSocket endpoint at brd.superproxy.io, port 9222 for Puppeteer and Playwright and 9515 for Selenium, with zone credentials and examples in JavaScript, Python and C#. Zones are created in the control panel or by API, and the free Bright Data MCP server connects MCP clients to browsing and extraction. SourceBright Data, docs.brightdata.com/products/scraping-browser/configuration and brightdata.com/ai/agent-browserread 2026-10-06
Testing, Debugging & Optimization	Partial
Live DevTools inspection of a session and a list of error codes support debugging, and the session prints its inspect URL as it runs. Bright Data names no test cases, harness or quality gate. SourceBright Data, docs.brightdata.com/products/scraping-browser/configurationread 2026-10-06
Browser & Computer Use	Full
Bright Data runs headful Chromium browsers in its cloud, which developers experience as headless through an API or MCP, and agents drive them to navigate, click, scroll and interact, with fingerprinting, CAPTCHA solving, retries and IP rotation handled inside the session. By its own count, Bright Data supports 1M+ concurrent sessions, with 400M+ actions performed daily. SourceBright Data, brightdata.com/ai/agent-browserread 2026-10-06

The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded

Product updates

October 15, 2026·Browser/computer usePartially Verified

Version 3.1.0 of the Bright Data MCP server turns every tool on by default, so the Agent Browser tools no longer need to be enabled separately. Results from the browser, search, scrape and dataset tools are now marked as untrusted web content, so the model treats text on a visited page as data rather than instructions.

Bears on: MCP / tool calling / API

View source

View all Bright Data Agent Browser updates →Tracked since Oct 2026 · Verified from public vendor sources

Pricing

Free tier (1 GB a month on the Browser API pricing page); pay as you go $8 a GB; Scale $499 a month with 71 GB; Enterprise custom

not captured

Free tier

What is public

Nothing captured: retrieved materials cover the product but no plan structure or rates; the badge reflects absence of retrieved pricing evidence rather than a confirmed sales-led model.

Variable cost rationale

Proxy backed browser infrastructure is conventionally metered by traffic and session volume, though no billing model was captured in retrieved materials.

Sales call required

Mixed (some tiers require a call)

Free / trial

Free tier with 1 GB a month and no credit card per the Browser API pricing page; Bright Data's llms.txt says the free tier does not cover the Browser API

Lowest paid plan

Pay as you go at $8 a GB, no monthly commitment

Key ambiguities

The two first party pages disagree on whether the free tier covers the Browser API, and on the starting rate ($8 a GB pay as you go against from $5 a GB).

Agentic Index verified 2026-09-27

Official pricing page

Related vendors

Alternatives to Bright Data Agent Browser

The closest documented capability profiles to Bright Data Agent Browser among browser and computer-use agents tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.

AgentQL5.5 / 14Adds documented Deployment & Data Residency and Triggers & Channel Coverage, among others
TinyFish7.5 / 14Adds documented Workflow Orchestration and Human Oversight & Guardrails, among others
Stagehand7.0 / 14Adds documented Workflow Orchestration and Human Oversight & Guardrails, among others
Notte9.5 / 14Adds documented Workflow Orchestration and Human Oversight & Guardrails, among others
AGI, Inc.6.0 / 14Adds documented Workflow Orchestration and Human Oversight & Guardrails, among others
Airtop9.0 / 14Adds documented Workflow Orchestration and Human Oversight & Guardrails, among others

Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded

Compare Bright Data Agent Browser with other vendors →Compare all browser and computer-use agents →See where Bright Data Agent Browser ranks →

Data confidence: high

Head to head

Bright Data Agent Browser vs Browserbase