[RFC]: vLLM-Omni Diffusion Module — Q2 2026 Roadmap

Motivation.

This roadmap consolidates open RFC items, and community priorities for the vllm-omni diffusion model module. Items overlapping with the Q2 Roadmap RFC #2136 are treated as highest urgency.

P0 must-have · P1 target Q2 · P2 stretch/Q3 · — open slot · (RFC needed) = no tracking issue yet, open one to claim

Want to contribute?

• See #3629 for open diffusion model recipe tasks.
• See #1217 for open diffusion feature tasks.
• See #1854 for open diffusion quantization tasks.
• See #1832 for diffusion testing tasks.

---

Proposed Change.

---

1. Diffusion User Interface & Guidance

Goal: Reduce the barrier to entry for diffusion models through recipes, guardrails, and auto-generated capability documentation.

• [P0] Diffusion Model Recipes — Publish end-to-end usage recipes covering recommended configurations, and performance tuning for popular diffusion models. RFCs: #1355, #1806, #2015; PRs:#3323, #3271, #3114
• [P0] Pipeline Parameter Validation — Add pre-flight checks that detect invalid parameter combinations at startup, e.g. incompatible parallelism modes or incorrect diffusion sampling params. RFCs: #2189,#2072, PRs: #2653
• [P0] Pipeline × Feature Support Registry — Build a structured compatibility registry that maps each pipeline to its supported acceleration features. (#2189)
• [P0] #1456 — Relevant high-priority community issue; scope follows the latest issue description.

---

2. Systematic Core Features

2.1 DiT Continuous Batching & Step-Level Scheduling

Goal: Enable fine-grained scheduling of diffusion denoising steps to improve throughput and support streaming job management.

• [P0] Abort Request — Support cancellation of in-flight diffusion inference requests. (#1769)
• [P0] Step-wise & Continuous Batching — Decompose the denoising loop into step-level scheduling units, allowing continuous batching for diffusion models. RFCs: #874, #2063, PRs: #2652, #2729
• [P1] Data Parallelism for Diffusion — Add data-parallel support so that multiple requests can be dispatched to independent DP replicas concurrently. (#1706)

2.2 Reinforcement Learning Support

Goal: Enable diffusion models to serve as generative policies within RL rollout.

• [P0] RL Inference Interface — Expose inference primitives required by RL training, including log-probability computation and gradient passthrough. (#2357, #778)
• [P1] LoRA Enhancements — LoRA loading acceleration, multi-LoRA switching, and non-PEFT format support. RFCs: #2198, #2150, #2149, #1396, #2309, PRs:#2783
• [P1] Diffusers Backend — Instead of vLLM-Omni, using diffusers as the inference backend. RFCs: #1456, PRs: #2724

2.3 Multi-Stage Orchestration

Goal: Allocate GPU/CPU resources across pipeline stages (text encoder → DiT → VAE) to accelerate diffusion models.

• [P1] Cross-Stage Orchestration & Resource Allocation — Optimized stage deployment for models like Bagel that benefit from cross-stage resource splitting. RFCs: #2089 PRs: #3208

---

3. World Model Support

Tracking: #1987

Goal: Provide the infrastructure needed for world models (autoregressive diffusion, VLM, 3D, etc.) to support various features.

• [P0] World Model Support — Support the inference acceleration of World Models, such as interactive video generation model, VLA(Vision-Language-Action) model. For example, DreamZero RFCs: #1987
• [P1] PagedAttention & KV Cache for Autoregressive Diffusion — Wire KV cache reuse into AR-diffusion inference to avoid full recomputation over growing context windows. RFCs: #2305, PRs: #2949
• [P1] Semi-Causal Attention Support — Support flexible switching between autoregressive and full attention; requires attention backend changes. PRs: #1975 , #2981
• [P1] Pipeline Parallelism & Stream Batch — Introduce pipeline parallelism and stream batching for real-time video generation, like StreamDiffusionV2. RFCs: #2280 PRs: #2322, #3099
• [P1] Streaming Inputs & Outputs — Frame-level streaming generation, instantaneous condition insertion, and checkpoint/resume for long-horizon inference. RFCs: #2201, PRs:#2342, #3157

---

4. Model Coverage Expansion

Goal: Merge all in-progress model PRs and systematically fill feature-coverage gaps in the existing model matrix.

• [P0] In-Progress Model PRs — Complete and merge six actively developed model-support PRs. Full test coverage required before merge. (#877, #1759, #1866, #2015, #2181, #1806)
• [P0] Quantization Coverage Expansion — Extend FP8/Int8 quantization support to additional image and video generation models. (#1854)
• [P1] Feature Matrix Completion — Systematically extend acceleration features to models with incomplete coverage; run feature compatibility tests. (#1217, #2083)

---

5. Online Serving & Entrypoint

Goal: Complete missing API endpoints, harden existing ones, and improve production reliability for long-running generation jobs.

• [P0] Text-to-Audio Endpoint — Deliver a /v1/audio/generations endpoint for audio diffusion models (e.g., Stable-Audio-Open) aligned with the OpenAI Audio API. (#1794)

---

6. Testing, CI/CD & Documentation

Goal: Establish continuous end-to-end quality assurance and keep documentation in sync with the codebase.

• [P0] Continuous L4 E2E Functional Testing — Stand up CI tests for diffusion models (potentially using smaller random-weight models). (#1832)
• [P0] Performance & Accuracy Benchmark Testing — Performance regression and accuracy monitoring tests in CI. (#2185) (RFC needed)
• [P1] L5 Testing — Test the stability and reliabilty for long-term severing (RFC needed)
• [P1] Docs Refinement — Overhaul offline inference and online serving documentation with complete, runnable examples across multiple settings. (#1244)

---

CC List.

cc @hsliuustc0106 @Yikun @wangxiyuan @Fishermanykx @jiangmengyu18 @blian6 @SamitHuang @ZJY0516 @linyueqian @david6666666 @Gaohan123 @tzhouam @princepride @fhfuih

Any Other Things.

No response

Before submitting a new issue...

• Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.