Matrix-Game 2.0 vs DreamX-World-5B: inputs, controls and API cost
Compare Matrix-Game 2.0 and DreamX-World-5B APIs: inputs, outputs, streaming delivery and ready-second cost. Choose a Windflow model for your product.
Matrix-Game 2.0 and DreamX-World-5B are both navigable world model options in Windflow's public catalog. Choose by the inputs and interaction you need, then evaluate quality and responsiveness with your actual application. This page compares the published serving profiles; it does not present a head-to-head performance experiment.
Quick decision
- Evaluate Matrix-Game 2.0 for: Image-seeded explorable worlds controlled purely by movement.
- Evaluate DreamX-World-5B for: Portrait-format interactive worlds for mobile-first experiences.
Specifications side by side
| Published specification | Matrix-Game 2.0 | DreamX-World-5B |
|---|---|---|
| Architecture | Real-time AR world model (SkyReels-V2 base, Self-Forcing distilled, 3-step) | 5B controllable world generation model (Wan2.2-TI2V backbone) |
| Resolution | 640 × 352 | 704 × 1280 portrait in the production runtime; negotiated per session |
| Chunk | 12-frame AR blocks; actions land at the next block | Action-sequence conditioned rollout |
| Control latency | One block from input to pixels | Not published |
| Inputs | A seed image and persistent WASD/look controls (no text prompt) | A seed image, a text prompt, and movement/camera-control state |
| Outputs | A navigable generated world stream on main_video | A controllable world rollout on main_video |
| Delivery | WebSocket (H.264) · WebRTC | WebSocket (H.264) · WebRTC |
| Session protocol | sdk.realtime.v2 | sdk.realtime.v2 |
| Price | $0.0024 per ready second | $0.0023 per ready second |
| Implementation status | production | production |
Implementation status describes runtime maturity. Live GPU readiness is reported separately; Windflow is in private beta and brings capacity up on demand.
Input and control fit
DreamX-World-5B accepts text prompts; Matrix-Game 2.0 does not in its published profile. Both models accept reference or seed images in the published Windflow profile. Both models do not accept uploaded source clips in the published Windflow profile. Both models do not accept live webcam input in the published Windflow profile. Both models accept movement and look controls in the published Windflow profile.
A shared session API does not make commands interchangeable. Use each model's supported inputs and control timing. In particular, prompt changes may land at a chunk or generation-window boundary; movement input can represent persistent state until replaced.
What Matrix-Game 2.0 is designed for
Matrix-Game 2.0 is Skywork AI's open-source, real-time interactive world model. It generates a navigable world from a single seed image and keyboard-style controls, with no text prompt. On Windflow it runs as a realtime session you steer with persistent movement and look.
- Image-seeded explorable worlds controlled purely by movement.
- Game-style demos where the player's input, not a prompt, drives the scene.
- Research and prototyping around action-conditioned video generation.
- Comparing action-conditioned world models on the same session protocol.
What DreamX-World-5B is designed for
DreamX-World is a general-purpose interactive world model from Alibaba's AMAP research team. DreamX-World-5B, built on a Wan2.2-TI2V backbone, is served on Windflow as a portrait-format controllable world rollout driven by a seed image, a text prompt and movement and camera-control state.
- Portrait-format interactive worlds for mobile-first experiences.
- Prompt-and-action controlled scene rollouts from a single image.
- Short controllable scenes for storyboarding and previsualisation.
- Evaluating a 5B world model against larger alternatives.
Compare a realistic session budget
- Matrix-Game 2.0: $0.0024 per ready second, or $1.44 for ten ready minutes.
- DreamX-World-5B: $0.0023 per ready second, or $1.38 for ten ready minutes.
Queueing and worker startup are free. A ready owner session can remain billable while delivery is paused or unwatched; release it when done. Independent sessions carry independent meters. Use the cost calculator to estimate your duration and daily sessions.
Evaluate before choosing
Use one repeatable input sequence, the same client and network, and equivalent output expectations. Record first visible frame, control-to-visible delay, frame gaps, scene continuity and rejected controls. Run several starts and a sustained session. If serving profiles differ, label the differences instead of treating displayed FPS as directly comparable native generation speed.
See the latency measurement guide and the model selector for a reproducible decision process.
Sources
- SkyworkAI/Matrix-Game on GitHub — MIT
- AMAP-ML/DreamX-World on GitHub — Apache-2.0
FAQ
Is Matrix-Game 2.0 better than DreamX-World-5B?
The answer depends on the inputs, control behavior, quality and output your product requires. The published specifications identify capability differences; an application-level evaluation establishes performance for your scenario.
Can I switch between these models through the same API?
Windflow keeps the session lifecycle common. Your application must adapt to each model's supported inputs, commands, output tracks and generation timing.
Do the published prices guarantee available GPU capacity?
No. Rates and implementation status are catalog facts. Live readiness is a separate operational signal, and private-beta access is required to run sessions.