Browser Use vs Project Mariner

Browser Use vs Project Mariner: side-by-side comparison of two computer-use & desktop agents systems — architecture, taxonomy, license, pricing, MCP/A2A support, and direct edges.

Browser Use · Project Mariner

Cost & capability

Browser UseProject Mariner
Capability bandcompetentcompetent
Capability composite6873
Cost tierfreemid
$/Mtok input0
$/Mtok output0
Use casesScoped AgenticScoped Agentic

Where they differ (14)

Rows where both sides have data and the values disagree — the shortlist of dimensions that actually distinguish these two systems.

Browser UseProject Mariner
Capability composite6873
Cost tierfreemid
TypeOSS browser-agent library + Cloud (browser-use)Google DeepMind browser-agent (research preview)
Created2024-10 (GitHub first commit)2024-12
PricingOSS free; Cloud tiers (usage-based)Bundled with Gemini Advanced ($19.99/mo)
Funding$17M seed Mar-2025 (Felicis lead)Google DeepMind parent (Alphabet)
Backend storageLocal Playwright state; Cloud uses managed PostgresGoogle-managed session state
DeploymentPip self-host + Browser Use CloudChrome extension + Google Cloud
API surfacePython SDK + Cloud REST APIChrome extension; underlying Gemini 2.0 API
Multi-tenancyCloud is multi-tenant; OSS is single-tenant per processSingle-tenant per Google account
MCPBrowser Use MCP server (community + first-party)searched not found
A2Asearched not foundGoogle's own A2A protocol — likely first-party compatible
Optimised forLLM-driven browser automation as a primitive — DOM + screenshot modeGemini 2.0-powered Chrome agent
Anti-fitNot for desktop apps (browser-only); not a SaaS — caller builds product on topResearch preview only; Chrome only; Gemini Advanced tier

At a glance

Browser UseProject Mariner
SectionComputer-use & desktop agents Computer-use & desktop agents
TierT1 T1
TypeOSS browser-agent library + Cloud (browser-use) Google DeepMind browser-agent (research preview)
Created2024-10 (GitHub first commit) 2024-12
Latest releaseActive weekly releases (PyPI: browser-use)
LicenseMIT
GitHubgithub.com/browser-use/browser-use — 60k+ stars
PricingOSS free; Cloud tiers (usage-based) Bundled with Gemini Advanced ($19.99/mo)
Funding$17M seed Mar-2025 (Felicis lead) Google DeepMind parent (Alphabet)
Backend storageLocal Playwright state; Cloud uses managed Postgres Google-managed session state
DeploymentPip self-host + Browser Use Cloud Chrome extension + Google Cloud
API surfacePython SDK + Cloud REST API Chrome extension; underlying Gemini 2.0 API
Multi-tenancyCloud is multi-tenant; OSS is single-tenant per process Single-tenant per Google account
MCPBrowser Use MCP server (community + first-party) searched not found
A2Asearched not found Google's own A2A protocol — likely first-party compatible
OpenTelemetrysearched not found searched not found
Optimised forLLM-driven browser automation as a primitive — DOM + screenshot mode Gemini 2.0-powered Chrome agent
Anti-fitNot for desktop apps (browser-only); not a SaaS — caller builds product on top Research preview only; Chrome only; Gemini Advanced tier

Taxonomy

AxisBrowser UseProject Mariner
storagekvkv
retrievalinjectioninjection
persistencesessionsession
updateagent-controlledagent-controlled
unitturnturn
governanceinspectableopaque
conflictstatelessstateless

Pros & cons

Browser Use

Pros: Most-starred browser-agent OSS in 2025; permissive MIT license; Felicis-backed; works on top of any major LLM.

Cons: Not a finished app — caller builds product on top; compliance / SOC 2 still maturing on Cloud.

Project Mariner

Pros: Strong WebVoyager numbers; Google distribution; A2A first-party adjacency.

Cons: Research preview; Chrome / Gemini Advanced tier; closed source.

Rows last verified 2026-05-14 / 2026-05-14. Data is CC-BY-4.0 — see how to read this.