# Hy-Nomas API

> Hy-Nomas (Distilled Hy-Nomas DiT with a stable-prefix PixelUMM decoder) as a hosted streaming API. Silent generated video on main_video. $0.0047 per billable second.

Source: https://windflow.online/models/hy-nomas · Navigable world model · Windflow Labs

## What it is

Hy-Nomas is an open interactive world model from Tencent Hunyuan. Describe a scene in text and Windflow streams a world from it at 832 × 480 and 24 fps, two seconds at a time: held keys move and turn the view, and text events change what happens next without restarting the world.

## Specifications

- **Architecture:** Distilled Hy-Nomas DiT with a stable-prefix PixelUMM decoder
- **Resolution:** 832 × 480
- **Playout:** 24 fps delivery target
- **Chunk:** 48 requested frames; variable committed prefix and a finalized tail
- **Control latency:** Prompt and camera changes take effect at the next segment boundary, a few seconds after they are sent
- **Inputs:** Text and persistent camera movement/look snapshots
- **Outputs:** Silent generated video on main_video
- **Delivery:** WebSocket (H.264) · WebRTC
- **Session protocol:** sdk.realtime.v2
- **Price:** $0.0047 per billable second

## Pricing

$0.0047 per billable session-second ($17.00/hr of billable time). Billing starts at the session's first generation_started event and stops when the session is released or ends. Queueing, worker startup, time before generation starts and the documented reconnect grace are not billed. After generation starts, paused or unwatched time remains billable; release the owner session to stop the meter.

## Use cases

- Explorable scenes opened from a single text description.
- Interactive stories where each turn adds an event to the same world.
- Walking a location with the keyboard before committing to a shot.
- Comparing text-opened and image-seeded world models on one protocol.

## Where it comes from

- [Tencent-Hunyuan/Hy-Nomas on GitHub](https://github.com/Tencent-Hunyuan/Hy-Nomas) — Apache-2.0
- [tencent/Hy-Nomas on Hugging Face](https://huggingface.co/tencent/Hy-Nomas) — Apache-2.0

## FAQ

### Does Hy-Nomas need a seed image?

No. A text prompt opens the world; held keys and later prompts steer it.

### How do controls reach the model?

Keys and text events are sampled at the start of each two-second segment, so a change shows from the next segment on.

### How long is one world?

Up to 60 seconds, 1,440 frames at 24 fps. Open the next world with reset or a new session.

### How much does Hy-Nomas cost on Windflow?

Hy-Nomas costs $0.0047 per billable session-second ($17.00/hr of billable time). Billing starts at the session's first generation_started event and stops when the session is released or ends. Queueing, worker startup, time before generation starts and the documented reconnect grace are not billed. After generation starts, paused or unwatched time remains billable; release the owner session to stop the meter. There is no subscription.

### How do I get access to Hy-Nomas?

Windflow is in private beta. Request access from the sign-up page, or ask for a live demo of Hy-Nomas. Approved accounts start with $5 and four concurrent sessions.

### Is Hy-Nomas always running?

No. Model capacity is brought up on demand rather than kept standing. Live readiness for each model is published at api.windflow.online/v1/status, and a live demo can be arranged on request.

## Related guides

- [How to use the Hy-Nomas API](https://windflow.online/guides/hy-nomas-api)