🌐 The open-source Agentic browser; alternative to ChatGPT Atlas, Perplexity Comet, Dia.
11.8k
Stars
1.2k
Forks
34
Open issues
13
Contributors
AI Analysis
BrowserOS is an open-source Chromium-based browser that natively runs AI agents for browser automation with 53+ tools, enabling tasks like navigation, clicking, and data extraction via natural language. It is purpose-built for users seeking privacy-first agentic browsing with support for local models (Ollama, LM Studio) or external APIs (Claude, OpenAI, Gemini), complemented by MCP server integration, scheduled task automation, and 40+ app integrations. This tool serves developers, researcher...
Inferred from signals mentioned in the README (tests, CI, type safety) — not a review of the actual code.
AI's overall editorial judgment — not an average of the bars above, can weigh other factors too.
BrowserOS: Open-source Chromium fork with native AI agents, targeting privacy-first browser automation
BrowserOS is a Chromium-based browser that embeds AI agent capabilities directly into the browser runtime, rather than layering automation on top of an existing browser. It targets developers, power users, and researchers who want browser automation via natural language but are unwilling to route data through cloud-controlled products like ChatGPT Atlas or Perplexity Comet. Key differentiators include BYOK (bring-your-own-key) support, local model compatibility via Ollama, an MCP server for programmatic control, workflow automation with a visual graph builder, and 40+ app integrations. The AGPL-3.0 license enforces open-source compliance for derivative works.
Created May 2025, positioning itself against a wave of AI-native browser announcements from large commercial players. Reached ~11.5K stars within roughly 13 months of creation, suggesting launch momentum driven by the AI browser trend.
Growth appears to have been driven by launch-period social media coverage and positioning against high-profile commercial alternatives (ChatGPT Atlas, Perplexity Comet, Dia). The 37 stars gained in the last 7 days as of June 2026 indicates growth has normalized significantly from any initial spike — the project is in a steady, slower accumulation phase rather than a viral growth phase. Community channels (Discord, Slack) and active documentation suggest intentional community-building rather than purely passive star accumulation.
Adoption not verified via third-party sources. Binary downloads are available for macOS, Windows, Linux (AppImage and Debian), which lowers the barrier to real-world use. Discord and Slack communities exist but community sizes are not disclosed. Star count (11.5K) and fork count (1,174) suggest meaningful developer interest, but these are not direct evidence of active production deployments. No public case studies, enterprise references, or download statistics are cited in the README.
Likely a Chromium fork with a TypeScript-based UI/agent layer built on top of the browser shell. The README references 53+ browser automation tools, an MCP server, a visual workflow graph builder, and persistent memory — suggesting a multi-layered architecture combining a modified browser binary with an embedded AI orchestration runtime. The use of TypeScript as primary language suggests the agent/UI layer is JS/TS while the underlying Chromium code would be C++. Exact separation between forked browser code and added agent layer is not documented in README.
Not documented in README.
Last push was June 22, 2026 — same day as analysis date — indicating active, recent development. The project is under 14 months old with consistent recent activity. Feature table in README is broad and detailed, suggesting ongoing development rather than a stale showcase project. Presence of a dedicated docs site (docs.browseros.com) and multiple community channels further supports active maintenance.
ADOPT IF: you need browser automation with strong privacy guarantees, want to use local LLMs or your own API keys without cloud intermediaries, and are comfortable with a young project that may have rough edges as a Chromium fork. AVOID IF: you require enterprise-grade stability, long-term vendor support, or need guarantees that Chromium security patches are applied promptly — Chromium forks carry real maintenance burden and the team's capacity to track upstream is unproven at this age. MONITOR IF: you are building developer tooling or AI agent pipelines and want to track whether the MCP server and workflow builder mature into reliable, production-grade components over the next 6-12 months.
Independent dimensions
Mainstream potential
4/10
Technical importance
7/10
Adoption evidence
2/10
- Chromium fork maintenance burden is substantial — tracking upstream security patches is a continuous, resource-intensive obligation that small teams often struggle to sustain long-term.
- The project is under 14 months old; core features like scheduling, memory, and workflow builder may still be in early/beta states with undocumented instability.
- AGPL-3.0 license may limit adoption in enterprise or commercial contexts where open-source copyleft is a legal concern.
- The competitive landscape is crowded and fast-moving — larger projects (vercel-labs/agent-browser, browser-use) and well-funded commercial alternatives may absorb the user base before BrowserOS reaches critical adoption mass.
- No verifiable production adoption data is publicly available; community size and actual active user count are unknown, making it difficult to assess real-world reliability feedback.
BrowserOS is likely to remain a relevant option in the open-source AI browser niche over the next 12-18 months if the team sustains its current development pace. Mainstream dominance appears unlikely given stronger-starred competitors and commercial incumbents, but it may carve out a durable position among privacy-focused developers and local AI enthusiasts.
Newsletter
Get analyses like this every Monday
Free weekly digest of the most interesting open-source discoveries.
Languages
Information
- Website
- https://BrowserOS.com
- Language
- TypeScript
- License
- AGPL-3.0
- Last updated
- 8h ago
- Created
- 14mo ago
- Analyzed with
- anthropic/claude-haiku-4-5
Stars over time
Contributors over time
Top 100 contributors only — repos with more will plateau at 100.
Open issues
[BUG] New tab page chat box does not respect Agent mode switch - always defaults to Chat mode
take_enhanced_snapshot missing
[BUG] "Show in folder" (folder icon) does nothing on macOS, even when the file still exists
[Feature Request]
Browser/Agent crash during mid/long-running tasks
Open pull requests
fix: replace Record<string, any> with Record<string, unknown> in GraphQL hooks
chore(deps)(deps-dev): bump @typescript/native-preview from 7.0.0-dev.20260622.1 to 7.0.0-dev.20260628.1 in /packages/browseros-agent
feat(agent): adaptive layout for control-tab previews
Top contributors
Similar repos
ntegrals/openbrowser
Open Browser is a TypeScript framework that enables AI agents to autonomously...
browser-use/browser-use
Browser Use is a Python library that enables AI agents to control a real web...
| Repository | Stars | Week Δ | Language | Score | Updated |
|---|---|---|---|---|---|
|
browseros-ai/BrowserOS
this
|
11.8k | +161 | TypeScript | 8/10 | 8h ago |
|
ntegrals/openbrowser
|
9.5k | — | TypeScript | 8/10 | 3mo ago |
 platonai/Browser4
|
1.1k | — | Kotlin | 7/10 | 9h ago |
|
browser-use/browser-use
|
104.1k | — | Python | 9/10 | 6h ago |
 epiral/bb-browser
|
6k | — | TypeScript | 8/10 | 1mo ago |
 vercel-labs/agent-browser
|
38.2k | — | Rust | 7/10 | 2d ago |
browser-use (100K stars, Python) is a library for controlling existing browsers programmatically — not a browser itself. It targets developers embedding automation in Python pipelines. BrowserOS targets end-users and developers who want a complete browser experience with AI built in. They are complementary more than directly competing, but both compete for mindshare in the 'AI browser automation' space.
nanobrowser (13.3K stars, TypeScript) is the closest structural competitor — also a TypeScript-based open-source AI browser project. BrowserOS has fewer stars but a more recent creation date and appears to offer a broader feature set (workflows, scheduling, MCP server, app integrations). Relative maturity and community size are similar.
openbrowser (9.5K stars, TypeScript) appears to be in the same category. BrowserOS has more stars but the comparison is close enough that neither has a decisive community lead. Differentiation likely comes down to feature set specifics not fully resolvable from metadata alone.
agent-browser (36.7K stars, Rust) has significantly more stars and is backed by Vercel's ecosystem credibility. Its Rust implementation may offer different performance tradeoffs. The Vercel association likely provides stronger integration with web development tooling, targeting a different primary audience than BrowserOS's privacy-first BYOK positioning.
BrowserOS explicitly positions against these closed commercial products. The differentiator is privacy (no data leaves the machine), cost control (BYOK), and extensibility (open source, MCP, local models). Commercial products may offer better polish and integrated AI capabilities out of the box, but require trusting vendor data handling.
