> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bytespike.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Veo 3.1 Fast

> Veo 3.1 tuned for ~30s turnaround — lower fidelity, faster iteration on natural-world footage.

**Vendor:** Google
**Model ID:** `veo-3.1-fast`
**Capability:** 720p · up to 5s · text + image init · async via tasks API
**Pricing:** per second, fast tier ([live rate](https://bytespike.ai/pricing#video))

Veo 3.1 Fast is the iteration variant. Same Veo motion strengths
(natural-world footage, coherent physics) capped at 720p / 5s, with
turnaround typically under 30 seconds. Right pick for prompt design,
draft passes before committing to Veo 3.1, and any UX where the user
clicks "generate" and is watching the clock.

## Submit

```bash theme={null}
curl https://llm.bytespike.ai/v1/tasks/submit \
  -H "x-api-key: $BYTESPIKE_API_KEY" \
  -H "content-type: application/json" \
  -d '{
    "model": "veo-3.1-fast",
    "prompt": "Slow zoom into a bowl of ramen, steam rising.",
    "duration_seconds": 4,
    "resolution": "720p"
  }'
```

### Body parameters

| Field              | Type    | Required | Default | Notes                    |
| ------------------ | ------- | -------- | ------- | ------------------------ |
| `model`            | string  | yes      | —       | `veo-3.1-fast`           |
| `prompt`           | string  | yes      | —       | English-tuned.           |
| `duration_seconds` | integer | yes      | —       | 1–5.                     |
| `resolution`       | string  | no       | `720p`  | Only `720p` supported.   |
| `aspect_ratio`     | string  | no       | `16:9`  | `16:9` / `9:16` / `1:1`. |
| `seed`             | integer | no       | —       | Reproducibility.         |
| `image_init`       | string  | no       | —       | URL of init image.       |

## Submit + poll

```json theme={null}
// Submit response
{"task_id": "task_…", "status": "queued", "estimated_credits": 0.18}

// Completed response
{"task_id": "task_…", "status": "completed", "result": {"video_url": "https://cdn.bytespike.ai/vid/...", "duration_seconds": 4, "resolution": "720p"}, "credits": 0.18}
```

Recommended cadence: 1s for first 30s, then 2s up to 60s. Query is free.

## Code examples

<CodeGroup>
  ```bash cURL theme={null}
  curl https://llm.bytespike.ai/v1/tasks/submit \
    -H "x-api-key: $BYTESPIKE_API_KEY" \
    -H "content-type: application/json" \
    -d '{"model": "veo-3.1-fast", "prompt": "Slow zoom into ramen", "duration_seconds": 4, "resolution": "720p"}'
  ```

  ```python Python theme={null}
  import time, requests

  API = "https://llm.bytespike.ai/v1"
  HEADERS = {"x-api-key": "$BYTESPIKE_API_KEY"}

  submit = requests.post(
      f"{API}/tasks/submit", headers=HEADERS,
      json={"model": "veo-3.1-fast", "prompt": "Slow zoom into ramen", "duration_seconds": 4, "resolution": "720p"},
  ).json()
  task_id = submit["task_id"]

  while True:
      r = requests.get(f"{API}/tasks/query", params={"task_id": task_id}, headers=HEADERS).json()
      if r["status"] == "completed":
          print(r["result"]["video_url"]); break
      if r["status"] in ("failed", "cancelled"):
          raise RuntimeError(r)
      time.sleep(1)
  ```

  ```javascript Node theme={null}
  const API = "https://llm.bytespike.ai/v1"
  const headers = { "x-api-key": process.env.BYTESPIKE_API_KEY, "content-type": "application/json" }

  const { task_id } = await fetch(`${API}/tasks/submit`, {
    method: "POST", headers,
    body: JSON.stringify({ model: "veo-3.1-fast", prompt: "Slow zoom into ramen", duration_seconds: 4, resolution: "720p" }),
  }).then((r) => r.json())

  while (true) {
    const r = await fetch(`${API}/tasks/query?task_id=${task_id}`, { headers }).then((r) => r.json())
    if (r.status === "completed") { console.log(r.result.video_url); break }
    if (r.status === "failed" || r.status === "cancelled") throw new Error(JSON.stringify(r))
    await new Promise((r) => setTimeout(r, 1000))
  }
  ```
</CodeGroup>

## Errors

| Code                  | Trigger                           | Billed?         |
| --------------------- | --------------------------------- | --------------- |
| 400 / 401 / 402 / 403 | Standard                          | No              |
| 451                   | Prompt blocked by upstream safety | No              |
| 5xx                   | Upstream issue                    | No (auto-retry) |

## When to use

* Latency-sensitive UX where the user is watching the clock.
* Prompt iteration before committing to a Veo 3.1 final render.
* For full-fidelity 1080p, see [Veo 3.1](/api-reference/video/veo-3-1).

## Limits

| Limit                       | Value           |
| --------------------------- | --------------- |
| Max duration                | 5s              |
| Min duration                | 1s              |
| Resolutions                 | 720p only       |
| Aspect ratios               | 16:9, 9:16, 1:1 |
| Supports image init         | Yes             |
| Typical latency for 4s clip | 20-40s          |
| Async via tasks API         | Yes             |