# Flux 3 | Image to Video Timestamped

FLUX.3 is Black Forest Labs’ frontier keyframe-to-video model, generating smooth, coherent motion between multiple keyframes for precise control over camera movement, timing, and scene progression.

## API Information

- **Model Slug:** flux-3-image-to-video-timestamped
- **Branded URL:** https://www.eachlabs.ai/black-forest-labs/flux-3/flux-3-image-to-video-timestamped
- **Provider:** Black Forest Labs
- **Category:** Image to Video
- **Output Type:** video
- **Status:** active
- **Version:** 0.0.1
- **Base Cost:** FLUX 3 Video image-to-video (timestamped keyframes) hd (default): $0.17 per second of output video, partial seconds rounded up.
- **Estimated Processing Time:** 260 seconds
- **Last Updated:** 2026-08-04
- **Interactive Demo:** https://www.eachlabs.ai/ai-models/flux-3-image-to-video-timestamped

## Pricing

- **Charge Type:** dynamic
- **Pricing Details:** FLUX 3 Video image-to-video (timestamped keyframes) hd (default): $0.17 per second of output video, partial seconds rounded up.

### Pricing Rules

| Condition | Pricing |
| --- | --- |
| draft == "true" | FLUX 3 Video Draft image-to-video (timestamped keyframes) (hd only): $0.06 per second of output video, partial seconds rounded up. |
| resolution eq_i "fhd" | FLUX 3 Video image-to-video (timestamped keyframes) fhd: $0.29 per second of output video, partial seconds rounded up. |
| Rule 3 | FLUX 3 Video image-to-video (timestamped keyframes) hd (default): $0.17 per second of output video, partial seconds rounded up. |

## Input Schema

No input parameters documented.

## Example Request

```bash
curl -X POST https://api.eachlabs.ai/v1/prediction/ \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "flux-3-image-to-video-timestamped",
  "input": {}
}'
```

## Output Schema

Response returned by `GET /v1/prediction/{id}` when the job completes:

```json
{
  "status": "success",
  "predictionID": "string",
  "output": "string (URL of generated video)",
  "metrics": {
    "predict_time": "number (seconds)"
  }
}
```

## Polling

```bash
curl https://api.eachlabs.ai/v1/prediction/{PREDICTION_ID} \
  -H "Authorization: Bearer YOUR_API_KEY"
```

| Status | Meaning |
|--------|---------|
| `processing` | Still running — poll again |
| `success` | Done — read `output` |
| `error` | Failed — read `message` / `details` |

## Webhook (alternative to polling)

Pass `"webhook_url": "https://your.host/path"` in the create request. Eachlabs POSTs this payload when the job ends:

```json
{
  "exec_id": "prediction-uuid",
  "status": "succeeded",
  "output": "https://...",
  "error": ""
}
```

`status` is `"succeeded"` or `"failed"`. `exec_id` equals the `predictionID` from create. Return 2xx within 30 seconds.

## Errors

Error body: `{ "status": "error", "message": "...", "details": "..." }`

| Code | Meaning |
|------|---------|
| `400` | Invalid input |
| `401` | Missing / invalid `Authorization` bearer token |
| `404` | Unknown model or prediction id |
| `429` | Rate limit — 100 creates / min, 10 concurrent per key |
| `5xx` | Retry with backoff |

## Overview

**Flux 3 | Image to Video Timestamped Overview**

**Flux 3 | Image to Video Timestamped** is an image-to-video model from **Black Forest Labs** that turns one or more keyframes into smooth motion with explicit timing control. It is designed for creators who need predictable scene progression, camera movement, and transitions rather than loosely guided animation. In the FLUX 3 family, this variant stands out because it is timestamped: users can anchor what should happen at specific moments, which makes it especially useful for structured motion planning and multi-shot continuity. Compared with generic **Black Forest Labs image-to-video** tools, the main value is tighter control over when visual changes occur and how the clip evolves across time.

**Technical Specifications**

  - **Model type:** Image-to-video generation with timestamped keyframe control.

  - **Input:** One or more reference images or keyframes, plus a text prompt and timing guidance where supported.

  - **Output:** Generated video clip.

  - **Resolution support:** Not publicly confirmed in the provided sources for this exact model.

  - **Max duration:** FLUX 3 video is described as supporting clips of up to 20 seconds in the broader family, but the exact cap for this timestamped variant is not separately confirmed.

  - **Aspect ratios:** Wide aspect-ratio support is described for FLUX 3 video generally, but exact presets are not confirmed here.

  - **Processing time:** No verified average runtime was found in the supplied sources.

  - **Architecture:** FLUX 3 is described as a unified multimodal model trained across image, video, and audio; the timestamped variant appears to be a video-specialized interface on that system.

**Key Considerations**

**Flux 3 | Image to Video Timestamped** is best for projects that need controlled movement across a short scene, not for long-form filmmaking. The most important prerequisite is a clear keyframe plan: the model performs best when the visual beats, camera intent, and timing are defined in advance. It is a strong fit for product shots, animated concept art, and transitions between staged frames. If you only need a fast stylized clip, a simpler image-to-video workflow may be enough. If you need precise timing, this model is the better choice, especially when using the **Flux 3 | Image to Video Timestamped API** in a repeatable production pipeline.

**Tips and Tricks**

Write prompts as a sequence of visual events, not as a single vague description. Name the subject, the motion, the camera behavior, and the moment of change. When using **Flux 3 | Image to Video Timestamped**, keep timestamps easy to follow and avoid asking for too many scene changes in one short clip. Start with one anchor image, then describe what must remain stable and what should evolve.

Examples:

“At 0s the product sits centered on a clean studio table; at 2s the camera slowly pushes in; at 4s the light shifts to reveal metallic texture.”

“Use the first keyframe as the opening pose, then transition smoothly by 3 seconds to a wider angle with gentle left-to-right camera drift.”

“Maintain the character’s face and wardrobe while the background changes gradually from dusk to neon city.”

For best results, keep motion physically plausible, use concise language, and avoid conflicting instructions about camera direction or pacing.

**Capabilities**

  - Converts still keyframes into motion-driven video sequences.

  - Supports timestamped control for timed visual changes.

  - Helps maintain scene continuity across multiple moments in a clip.

  - Enables controlled camera movement such as pushes, pans, and drifts.

  - Useful for staged transitions between composition states.

  - Fits workflows that need predictable motion planning for short-form video.

  - Works well for stylized visuals where timing matters as much as appearance.

**Use Cases for Flux 3 | Image to Video Timestamped**

**Creators** can use **Flux 3 | Image to Video Timestamped** to animate concept art into a polished reveal sequence. For example: “At 0s show the hero standing still; at 3s the cloak moves in the wind; at 5s the camera circles slightly behind.” This uses timestamped motion control to make a planned reveal feel cinematic.

**Marketers** can turn product stills into timed launch clips. Example prompt: “Start with the bottle on white background, then at 2s add a slow zoom, and at 4s rotate the label toward the camera.” This uses the model’s keyframe-to-motion flow for precise product presentation.

**Designers** can prototype motion studies for UI, environments, or brand visuals. Example: “From the first keyframe, transition the layout into a wider scene by 4s while keeping the logo fixed.” This benefits from stable scene progression.

**Developers** integrating the **Black Forest Labs image-to-video** workflow can build repeatable timed generation pipelines for social clips or automated storyboards.

**Things to Be Aware Of**

Timestamped video generation works best when the prompts are internally consistent. If the instructions conflict, the model may smooth over the request instead of following every detail exactly. Very complex motion, rapid scene cuts, or many simultaneous changes can reduce coherence. Users also commonly over-specify camera behavior, which can make the result feel rigid. For production use, expect iterative prompting and test with short clips before scaling a workflow through the **Flux 3 | Image to Video Timestamped API**.

**Limitations**

**Flux 3 | Image to Video Timestamped** is optimized for short, controlled motion, not long narrative sequences. Publicly verified details for exact resolution, aspect-ratio presets, and average latency were not available in the supplied sources for this specific variant. Like most image-to-video systems, it may struggle with highly complex actions, exact physics, and dense scene changes. It is also less suitable when a user needs unrestricted editing across many shots rather than a timed single-clip workflow.

## Usage Notes

- API Base URL: `https://api.eachlabs.ai/v1`
- Authentication: send `Authorization: Bearer YOUR_API_KEY`. Generate a key from the Eachlabs dashboard at https://www.eachlabs.ai/dashboard/api-keys.
- File-typed parameters (`*_url`, `image_url`, `video_url`, `audio_url`, etc.) accept publicly-reachable HTTPS URLs only. Upload your asset first (GCS / S3 / your CDN) and pass the resulting URL. Data-URIs and localhost URLs are rejected.
- For structured parameters (arrays / objects) send real JSON values, not stringified payloads.
- Monetary values are reported in USD; per-token / per-megapixel rates may be billed in micro-cents internally.
- Prefer `webhook_url` over polling for long-running predictions — see the Webhook Callback section.