[ROADMAP] TeleFuser Q2 2026

[lzx1413]

Overall Goal

1. Improve open-source project infrastructure 2. Fix existing bugs 3. Support new feature integration, such as KV cache

🧠 New Model Integrations

• SwiftVR

New Feature Support

• Streaming Inference with KV Management
• Disaggregated Inference

⚡ Performance & Kernels

• NVF4 for worldmodel
• support SVDQuant

🔗 Distributed Inference

• Ulysses SP for lingbot-world

🌐 API & Service

• support multi-instance in server

📚 Documentation & Tooling

• support online document website
• regression testing for pipeline runs

Deployment

• make pip install package avaliable

Progress Tracking

Status: 🔜 Not Started

Area	Progress	Key PRs/Issues
New Models	⬜ 0/1
New Features	⬜ 0/2
Performance	⬜ 0/1
Distributed	⬜ 0/1
API & Service	⬜ 0/1
Docs & Tooling	⬜ 0/1

---

[ActivePeter]

Is there any difference between SGLang’s KV cache implementation and TeleFuser’s? Fluxon has been working on SGLang KV cache integration recently.