Roadmap¶

Doc Version: 2.0.0 Last Updated: 2026-09-06 Git 标签: v3.0.1 Author: Lincoln

This document outlines the future development plans and feature roadmap for the JAiRouter project.

Project Vision¶

JAiRouter aims to become the best open-source AI model service routing gateway, providing users with:

• Unified Access: One gateway to access all AI model services
• Intelligent Routing: Smart load balancing based on multiple strategies
• High Availability: Robust fault tolerance and recovery mechanisms
• Ease of Use: Simple configuration and comprehensive documentation
• High Performance: Support for large-scale concurrency and low-latency responses
• Observability: Comprehensive monitoring, logging, and distributed tracing

Current Version Status¶

✅ v3.0.1 (Current Stable)¶

Release Status: Released (2026-09-06) Git Tag: v3.0.1

Key Features (v2.9.6 → v3.0.1 evolution)¶

Statistics¶

• Test count: 3,243 (all green)
• Java source files: 750+
• Codebase: ~135k LOC

✅ v2.10.x - Web Console Refactor Series (Completed, 2026-09-06)¶

Features are complete but the console did not yet form a full system flow; one theme per version: - v2.10.0 Frontend foundation ✅ (released 2026-09-05): dead-code cleanup (13 .vue + common ≈13.5k lines) + PageSkeleton/StatCard/useChartTheme + Dashboard·Service·Rule pilot + ~20 pages tokenized + Element Plus dark-mode fixes + Layout/console cleanup - v2.10.1 Governance hub ✅ (released 2026-09-06): Dashboard v2 governance-chain panel + exception summary + Cross-link navigation + 7 pages onto PageSkeleton + tokenization wrap-up - v2.10.2 Config onboarding & capability pages ✅ (released 2026-09-06): response-cache management page (status/invalidation + tier-B runtime toggle + hit observability) + slow-query analysis page + CB reset / quota alerts + onboarding guide + backend cache status & deprecations + 45 permission codes - v2.10.3 Web bilingual edition ✅ (released 2026-09-06): full zh/en i18n + language switcher + Element Plus locale + date/number localization + dark-mode contrast cleanup - v2.10.4 Web experience wrap-up ✅ (released 2026-09-06): unified service-type display names + ECharts language/theme hot-switch + dead-code cleanup (utils/errorHandler, validators/*, etc.) + dark-contrast audit gate (36 pages ≥4.5) + backend-message doc note + dual-theme code highlighting

🎯 v3.0.x - Web Complete-Flow Series (In Progress)¶

v3.0 is the Web complete-flow milestone (public /api/v1 stays compatible, no breaking): - v3.0.1 Flow wiring ✅ (released 2026-09-06): Onboarding preselect loop (useRoutePreselect + 5 containers) + Dashboard governance entries + CB/LB reverse jumps + version-page inbound + legacy redirect cleanup + PageSkeleton across 18 pages - v3.0.2 API cleanup & permission closure: remove the 4 deprecated controllers and TokenUsageController dead methods (DTO/Registry/tests cleaned with them) + progressive URL-rule registration for permissions (each rule + 401/403 test) - v3.0.3 Acceptance & release: end-to-end journey acceptance + README screenshot refresh (bilingual × light/dark) + docs/plan sync + Release (tag/gh/docker, dual remotes) - After v3.0.x the Web milestone closes; high-availability foundation (multi-node/Redis distributed/config rollback) re-assessed after v3.0; semantic cache evaluation only, separate project; microservice migration separate

✅ v2.6.11 LTS (Long-Term Support)¶

Release Status: Released (2026-04-17) Maintenance Period: Until 2028-05

Completed Features¶

• ✅ Multi-tenancy support
• ✅ API Key authentication
• ✅ JWT Token support
• ✅ OAuth 2.0 integration
• ✅ Role-Based Access Control (RBAC)
• ✅ Request/response data obfuscation
• ✅ Security audit logging
• ✅ H2 embedded database
• ✅ PostgreSQL/MySQL support
• ✅ Redis cache integration
• ✅ Prometheus metrics collection
• ✅ Grafana dashboard templates
• ✅ Distributed tracing (Zipkin/OpenTelemetry)
• ✅ Complete Docker deployment
• ✅ Kubernetes deployment support

Code Quality¶

• ✅ Checkstyle code standards
• ✅ SpotBugs static analysis
• ✅ JaCoCo test coverage reports
• ✅ 700+ unit tests
• ✅ E2E integration tests

🎯 v2.7.x - Performance Optimization Series (Q2-Q3 2026)¶

Completed¶

Performance Improvements¶

• Route flow optimization: 20-50%
• Memory usage reduction: 15%
• Startup time reduction: 30%

⏸️ v3.0 - Microservices Architecture (Indefinitely Postponed)¶

Status: ⏸️ Postponed

The v3.0 microservices architecture transformation has been indefinitely postponed as the current monolithic architecture meets all requirements.

Feature Roadmap¶

Adapter Extensions¶

Supported ✅¶

Planned 📋¶

• 📋 Google Gemini (v2.8.x)
• 📋 Cohere API (v2.8.x)
• 📋 AWS Bedrock (v2.9.x)

Load Balancing Strategies¶

Implemented ✅¶

• ✅ Random
• ✅ Round Robin
• ✅ Weighted Round Robin
• ✅ Least Connections
• ✅ IP Hash
• ✅ Consistent Hash

Planned 📋¶

• 📋 Latency-based
• 📋 Cost-based
• 📋 Model Capability-based

Rate Limiting Algorithms¶

Implemented ✅¶

• ✅ Token Bucket
• ✅ Leaky Bucket
• ✅ Sliding Window
• ✅ Warm Up
• ✅ Adaptive Rate Limiting

Planned 📋¶

• 📋 Distributed Rate Limiting
• 📋 User-based Rate Limiting
• 📋 API Key-level Rate Limiting

Monitoring and Observability¶

Implemented ✅¶

• ✅ Prometheus metrics
• ✅ Grafana dashboards
• ✅ Health check endpoints
• ✅ Basic alert rules
• ✅ Distributed tracing (Zipkin/OpenTelemetry)
• ✅ Structured logging

Planned 📋¶

• 📋 Business metrics analysis
• 📋 Cost analysis
• 📋 Custom alert rules

Technical Architecture¶

Current Architecture (v2.7.x)¶

Monolithic App → Modular Design → Reactive Programming

Architecture Evolution¶

Tech Stack¶

Core Technology¶

• Backend: Spring Boot 3.5.5 + WebFlux (Reactive)
• Frontend: Vue 3 + TypeScript + Element Plus
• Database: H2 (embedded) + R2DBC
• Cache: Redis (optional)
• Monitoring: Prometheus + Grafana
• Tracing: OpenTelemetry + Zipkin

Data Storage¶

• Default: H2 embedded database
• Production: PostgreSQL / MySQL
• Cache: Redis (optional)

Performance Targets¶

Community Development¶

Current Status¶

• GitHub / Gitee dual platform hosting
• Complete user and API documentation
• Chinese/English documentation support
• MkDocs static website

Development Goals¶

• Active developer community
• Regular version releases
• More adapter support
• Complete plugin system

Release Cycle¶

Contributing¶

Code Contributions¶

• Feature development and bug fixes
• Performance optimization and refactoring
• Test case writing
• Code review

Documentation Contributions¶

• User documentation improvements
• API documentation updates
• Tutorial and example writing
• Multi-language translation

Feedback Channels¶

• GitHub Issues: https://github.com/Lincoln-cn/JAiRouter/issues
• GitHub Discussions: https://github.com/Lincoln-cn/JAiRouter/discussions
• Documentation Feedback: Submit via GitHub PR

Risks and Challenges¶

Technical Risks¶

• Performance bottlenecks: Performance challenges at scale
• Compatibility: Multi-version API compatibility maintenance
• Security: Security vulnerabilities and attack protection

Market Risks¶

• Increased competition: Competition from similar products
• Technology changes: Challenges from rapid AI technology development
• User needs: Rapidly changing user requirements

Mitigation Strategies¶

• Continuous technical innovation and optimization
• Active community building and maintenance
• Flexible product strategy adjustments
• Comprehensive quality assurance system

Summary¶

JAiRouter will continue to uphold the open-source spirit and is committed to providing users with the best AI model service routing solution. We welcome community participation and contributions to jointly promote the project's development and progress.

Recent Focus (2026)¶

1. ✅ Complete v2.7.x performance optimization series + RBAC permission control
2. ✅ Advance v2.8.x configuration management + adapters + plugin system/rule engine/resource pools
3. ✅ v2.9.x intelligent routing series (failover/tag routing/response cache P0)
4. ✅ v2.9.10 response cache P1 (streaming cache + invalidation API + rate-limit short-circuit)
5. ✅ v2.9.11 README/screenshot refresh + hallmark web-console design audit + frontend fixes
6. ✅ v2.10.x Web console refactor series completed (v2.10.0 foundation ✅ 2026-09-05; v2.10.1 governance hub ✅ / v2.10.2 config & capabilities ✅ / v2.10.3 Web bilingual edition ✅ / v2.10.4 Web experience wrap-up ✅ 2026-09-06)
7. 🎯 v3.0.x Web complete-flow series (v3.0.1 flow wiring ✅ 2026-09-06; v3.0.2 API cleanup & permission closure 📋; v3.0.3 acceptance & release 📋)
8. 📋 Semantic cache evaluation (vector-similarity reuse, separate project); high-availability foundation (multi-node/Redis) re-assessed after v3.0.x

Long-term Vision¶

1. Become the standard in AI model routing
2. Build a complete ecosystem
3. Achieve enterprise-level commercial success
4. Drive industry technology development

Last Updated: September 6, 2026

For any suggestions or ideas, feel free to communicate with us via GitHub Discussions.