Network packet processor with a friendly UI for circumventing Deep Packet Inspection (DPI) systems.
1.6k
Stars
66
Forks
2
Open issues
5
Contributors
AI Analysis
B4 is a Linux-based network packet processor that bypasses Deep Packet Inspection (DPI) systems using netfilter queue manipulation, paired with a web UI for easy configuration. It is purpose-built for users in censorship-restricted regions who need to unblock content and services, and is not a general-purpose networking tool—it serves a specific, geopolitically-motivated use case.
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.
Go-based DPI bypass tool with web UI, gaining traction among circumvention-focused Linux users
B4 is a Linux network packet processor written in Go that circumvents Deep Packet Inspection (DPI) systems via netfilter queue manipulation. Built for users in regions with heavy content filtering, it provides a web UI, CLI, Docker support, and SOCKS5 proxy. Created August 2025, it has accumulated ~1,500 stars and 111 stars in the last week, suggesting early-phase adoption momentum within a niche audience. The project targets system administrators, router users, and individuals needing DPI evasion rather than mainstream consumers.
B4 emerged in August 2025 as a newer entry in the DPI-bypass ecosystem, which includes older, larger projects like GoodbyeDPI (28k stars, 2015 origin) and Zapret (15k stars). Unlike some predecessors, B4 is written in Go and emphasizes a modern web UI and Docker support, positioning itself as more accessible to non-expert users.
The project gained 111 stars in 7 days (as of July 2026), indicating recent acceleration. The last push on July 2, 2026 suggests active maintenance. This growth trajectory is notable but remains modest compared to established competitors; early momentum may reflect fresh release publicity or regional adoption waves, or both. The presence of Russian and English documentation and a Telegram channel suggests intentional community-building in regions with DPI concerns.
Adoption not verified through the README. No user testimonials, deployment case studies, or organizational usage statements are included. The 111 stars in 7 days and 1,523 total stars suggest interest, but GitHub activity does not confirm production deployment at scale. The Telegram channel may indicate a user community, but its size and engagement are not documented. Real-world usage remains an open question.
Based on README, B4 uses Linux netfilter queue (NFQUEUE) to intercept and manipulate network packets. It appears to support multiple DPI evasion strategies, ASN scanning, custom payload capture, and domain-based rules. The web UI is built separately and served on port 7000. Multi-architecture builds (amd64, arm64, armv7) and Docker packaging indicate attention to deployment flexibility. Code quality and architectural rigor cannot be inferred from README alone.
Not documented in README. No mention of test suites, CI/CD coverage, or validation practices. This is a gap in transparency but not necessarily an indicator of poor quality.
Last push July 2, 2026 (2 days before analysis date) indicates active maintenance. The project is ~10 months old, too young to have a long maintenance history. Release versioning is present (v1.43.0 mentioned), suggesting iterative improvement. No evidence of stalled issues or abandoned pull requests in README.
ADOPT IF: you run Linux servers or routers in regions with DPI filtering, prefer web UI over CLI-only tools, and want an actively maintained Go-based alternative to C implementations. AVOID IF: you need proven, battle-tested maturity (older projects like GoodbyeDPI have longer histories), require commercial support, or operate outside Linux environments. MONITOR IF: you're deciding between B4 and established competitors—adoption evidence is still limited, and long-term maintenance trajectory is unproven.
Independent dimensions
Mainstream potential
3/10
Technical importance
6/10
Adoption evidence
3/10
- Early-stage project (10 months old as of July 2026) with unproven long-term maintenance; single primary maintainer (DanielLavrushin) increases bus-factor risk.
- Adoption not verified—star count is encouraging but does not confirm production use at scale; early enthusiasm may not translate to sustained deployment.
- Linux-only constraint (requires netfilter NFQUEUE support) limits addressable market compared to cross-platform tools.
- GPL-3.0 license may restrict use in proprietary or restrictive deployment contexts.
- Test coverage and security auditing not documented; DPI evasion involves network security—lack of transparency on testing is a concern.
B4 is likely to remain a solid option within the Linux DPI-bypass niche, gaining gradual adoption in regions with heavy filtering. Unlikely to overtake GoodbyeDPI or Zapret due to their head-start and established user bases, but may carve out a distinct segment among users prioritizing ease-of-use and modern UI over raw feature depth. Sustainability depends on maintainer commitment; if maintained actively, it will stabilize as a credible alternative; if abandoned, it will fade.
Newsletter
Get analyses like this every Monday
Free weekly digest of the most interesting open-source discoveries.
Languages
Information
- Language
- Go
- License
- GPL-3.0
- Last updated
- 3d ago
- Created
- 11mo 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
Open pull requests
Top contributors
Similar repos
| Repository | Stars | Week Δ | Language | Score | Updated |
|---|---|---|---|---|---|
|
|
1.6k | +188 | Go | 7/10 | 3d ago |
|
|
4.7k | — | C | 7/10 | 6d ago |
|
|
15.6k | — | C | 7/10 | 1mo ago |
|
|
28.5k | — | C | 7/10 | 6mo ago |
|
|
6.1k | — | Kotlin | 7/10 | 2w ago |
|
|
1.9k | — | Python | 7/10 | 2mo ago |
Older, larger, C-based. B4 is newer, in Go, with web UI and Docker. GoodbyeDPI likely has broader adoption; B4 targets users preferring managed interfaces.
Feature-rich C implementation with broader DPI patterns. B4 appears to offer simpler configuration and web UI in exchange, potentially less technical depth.
Mobile-focused (Kotlin). B4 targets Linux servers and routers, complementary rather than direct overlap.
Successor to Zapret. B4 is concurrent, not a replacement; different language and UX approach.
Detection-focused, not evasion-focused. B4 is orthogonal—evasion after detection is identified.