# LTX 2.3 Trainer (V2) - Keyframe Interpolation API

> Trains a LoRA that generates the motion between first and last keyframes, optionally with a middle one

Trains a LoRA for keyframe interpolation: at inference you supply first and last frames and the model generates the video in between. Enable `include_middle_keyframe` to train on first, middle and last frames; every validation sample must then provide `middle_image_url`. Data comes as a zip via `training_data_url`, with up to 2 validation entries.

- Model id: `ltx23-trainer-v2/interpolate`
- Made by: Lightricks
- Category: Training
- Output: file
- Price: $0.0024 / step (cheapest run $4.80 (4800 API credits))
- Run: `POST https://api.doitong.com/v1/run/ltx23-trainer-v2/interpolate`
- Quote: `POST https://api.doitong.com/v1/quote/ltx23-trainer-v2/interpolate`
- Docs page: https://doitong.com/developers/models/ltx23-trainer-v2/interpolate

## Pricing

Price: **$0.0024 / step**. Billed per training step (`number_of_steps`).

The full example input below costs $4.80 = 4800 API credits (2000 steps × 2.4 API credits).

Prices are the upstream price in USD, paid from the API balance in API credits (1000 API credits = 1 USD). `POST https://api.doitong.com/v1/quote/ltx23-trainer-v2/interpolate` with the same JSON body returns the exact price for that input; the quote is exactly what a run charges. Invalid input is never charged and failed or canceled requests are refunded automatically.

## Input parameters

| Parameter | Canonical name | Type | Required | Default | Allowed values / range | Description |
|---|---|---|---|---|---|---|
| `training_data_url` |  | string | yes |  |  | URL for a `.zip` containing training data, organized according to the selected training mode. Each media file may have an accompanying `.txt` caption with a matching base name. |
| `rank` |  | enum (integer) | no | `32` | `8`, `16`, `32`, `64`, `128` | LoRA adaptation rank, trading increased memory consumption for greater capacity as the setting rises. |
| `number_of_steps` |  | integer | no | `2000` | 100 – 10,000 | Total count of training steps. |
| `learning_rate` |  | number | no | `0.0002` | 0.000001 – 1 | Optimization learning rate. |
| `number_of_frames` |  | integer | no | `89` | 9 – 121 | Frame count for each training sample. Dividing the count by 8 must leave a remainder of 1; examples include 1, 9, 17, 25, 33, 41, 49, 57, 65, 73, 81, 89, 97. |
| `frame_rate` |  | integer | no | `24` | 8 – 60 | Training video's target frame rate in frames per second. The native rate for LTX-2.3 is 24. |
| `resolution` |  | enum (string) | no | `medium` | `low`, `medium`, `high` | Training resolution. Increasing the resolution raises memory requirements. |
| `aspect_ratio` |  | enum (string) | no | `1:1` | `16:9`, `1:1`, `9:16` | Training aspect ratio. |
| `trigger_phrase` |  | string | no | `` |  | LoRA style activation phrase, inserted at the start of captions during training. |
| `auto_scale_input` |  | boolean | no | `false` |  | Automatic video adjustment to match the requested FPS and frame count when enabled; image datasets remain unchanged. |
| `split_input_into_scenes` |  | boolean | no | `true` |  | Scene splitting for videos exceeding a duration threshold, enabled when true. |
| `split_input_duration_threshold` |  | number | no | `30` | 1 – 60 | Scene-splitting trigger expressed as a duration in seconds: only videos longer than the threshold are split. |
| `debug_dataset` |  | boolean | no | `false` |  | Preprocessing inspection mode for checking videos, images, and captions before a full training run. Enabling it makes the trainer return the processed training dataset in a downloadable archive for manual review. |
| `validation` |  | array of object {prompt, start_image_url, end_image_url, middle_image_url} | no | `[]` | ≤ 2 items | Validation data supplied as a list; an entry combines keyframe images with a prompt. |
| `validation[].prompt` |  | string | yes |  |  | Prompt used during validation. |
| `validation[].start_image_url` |  | string (image URL) | yes |  | ≥ 1 characters | Keyframe image URL for the initial frame. |
| `validation[].end_image_url` |  | string (image URL) | yes |  | ≥ 1 characters | Keyframe image URL for the final frame. |
| `validation[].middle_image_url` |  | string (image URL) \| null | no |  |  | Middle-keyframe image URL for first+middle+last interpolation. Runs with `include_middle_keyframe=true` require it; otherwise, it is optional. |
| `validation_negative_prompt` |  | string | no | `blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts.` |  | Negative prompt applied during validation, which uses a single-stage preview process. Because that process approximates distilled production inference with two stages, its guidance and preview quality differ from the final results. |
| `validation_number_of_frames` |  | integer | no | `89` | 9 – 121 | Frame count for each validation video. |
| `validation_frame_rate` |  | integer | no | `24` | 8 – 60 | Target frame rate for validation videos, in frames per second; LTX-2.3 runs natively at 24. |
| `validation_resolution` |  | enum (string) | no | `high` | `low`, `medium`, `high` | Resolution used during validation. |
| `validation_aspect_ratio` |  | enum (string) | no | `1:1` | `16:9`, `1:1`, `9:16` | Aspect ratio used during validation. |
| `stg_scale` |  | number | no | `1` | 0 – 3 | Spatio-Temporal Guidance (STG) scale, with 1.0 recommended and 0.0 turning STG off. |
| `include_middle_keyframe` |  | boolean | no | `false` |  | Middle-keyframe inclusion for video generation: true requires `middle_image_url` in every validation sample and retains the middle frame alongside the first and last, rather than using the first and last alone. |

## Example input

Required parameters only:

```json
{
  "training_data_url": "https://example.com/your-file.zip"
}
```

A fuller example:

```json
{
  "training_data_url": "https://example.com/your-file.zip"
}
```

## Run it

Set `DOITONG_API_KEY` to a key from https://doitong.com/developers/dashboard#keys. `?wait=20` returns the finished request in the same call when it is ready within 20 seconds; otherwise poll `GET /requests/{id}` (it also accepts `?wait=`), or pass `webhook_url`.

### curl

```bash
# Run the model; wait up to 20 s for the finished result in the same call
curl -X POST "https://api.doitong.com/v1/run/ltx23-trainer-v2/interpolate?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "training_data_url": "https://example.com/your-file.zip"
  }'

# Still "queued" or "running"? Poll with the id from the response:
curl "https://api.doitong.com/v1/requests/REQUEST_ID?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY"
```

### JavaScript

```javascript
const API = "https://api.doitong.com/v1";
const headers = {
  "Authorization": "Bearer " + process.env.DOITONG_API_KEY,
  "Content-Type": "application/json",
};

// Run the model; wait up to 20 s for the finished result in the same call
let res = await fetch(API + "/run/ltx23-trainer-v2/interpolate?wait=20", {
  method: "POST",
  headers,
  body: JSON.stringify({
    "training_data_url": "https://example.com/your-file.zip"
  }),
});
let request = await res.json();
if (!res.ok) throw new Error(request.error.code + ": " + request.error.message);

// Not finished yet: long-poll until it is
while (request.status === "queued" || request.status === "running") {
  res = await fetch(API + "/requests/" + request.id + "?wait=20", { headers });
  request = await res.json();
}
if (request.status !== "succeeded") throw new Error(JSON.stringify(request.error));
console.log(request.outputs.map((o) => o.url), request.cost.credits + " API credits ($" + request.cost.usd + ")");
```

### Python

```python
import os
import requests

API = "https://api.doitong.com/v1"
headers = {"Authorization": "Bearer " + os.environ["DOITONG_API_KEY"]}

payload = {
    "training_data_url": "https://example.com/your-file.zip",
}

# Run the model; wait up to 20 s for the finished result in the same call
r = requests.post(API + "/run/ltx23-trainer-v2/interpolate", params={"wait": 20}, json=payload, headers=headers)
request = r.json()
if not r.ok:
    raise RuntimeError(request["error"]["code"] + ": " + request["error"]["message"])

# Not finished yet: long-poll until it is
while request["status"] in ("queued", "running"):
    request = requests.get(API + "/requests/" + request["id"], params={"wait": 20}, headers=headers).json()
if request["status"] != "succeeded":
    raise RuntimeError(request["error"])
print([o["url"] for o in request["outputs"]], request["cost"]["credits"], "API credits", "$" + request["cost"]["usd"])
```

### MCP

The MCP server at `https://api.doitong.com/v1/mcp` exposes the tools `search_models`, `get_model`, `quote`, `run`, `get_request`, `cancel_request`, `list_requests`, `create_upload` and `get_account`. A `run` call for this model:

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "run",
    "arguments": {
      "model_id": "ltx23-trainer-v2/interpolate",
      "input": {
        "training_data_url": "https://example.com/your-file.zip"
      },
      "wait_seconds": 20
    }
  }
}
```

## Response

Every call returns a request object. `status` is one of `queued`, `running`, `succeeded`, `failed`, `canceled`. `outputs` lists the generated files with URLs on our storage (kept forever, or 7 days with `?storage=temporary`; `storage` says which); `output` is the model's own result object with the same URLs.

```json
{
  "id": "req_8k2m4q7x1z9c3v5b6n0p2r4t",
  "object": "request",
  "model": "ltx23-trainer-v2/interpolate",
  "status": "succeeded",
  "created_at": "2026-10-20T12:00:00Z",
  "started_at": "2026-10-20T12:00:02Z",
  "completed_at": "2026-10-20T12:00:09Z",
  "queue_position": null,
  "cost": {
    "credits": 4800,
    "usd": "4.80",
    "charged": 4800,
    "refunded": 0
  },
  "input": {
    "training_data_url": "https://example.com/your-file.zip"
  },
  "output": {
    "video": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_name": "0-output.jpeg",
      "file_size": 1843200
    },
    "audio": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_name": "0-output.jpeg",
      "file_size": 1843200
    },
    "lora_file": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_name": "0-output.jpeg",
      "file_size": 1843200
    },
    "config_file": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_name": "0-output.jpeg",
      "file_size": 1843200
    },
    "debug_dataset": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_name": "0-output.jpeg",
      "file_size": 1843200
    }
  },
  "outputs": [
    {
      "type": "file",
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_size": 1843200,
      "stored": true
    }
  ],
  "error": null,
  "metrics": {
    "inference_seconds": 6.8
  },
  "urls": {
    "self": "https://api.doitong.com/v1/requests/req_8k2m4q7x1z9c3v5b6n0p2r4t",
    "cancel": "https://api.doitong.com/v1/requests/req_8k2m4q7x1z9c3v5b6n0p2r4t/cancel"
  }
}
```

### Output fields (`output`)

| Field | Type | Description |
|---|---|---|
| `video` | object {url, content_type, file_name, file_size} \| null | Combined video preview from validation, when available for endpoints that produce video. |
| `audio` | object {url, content_type, file_name, file_size} \| null | Combined audio from validation, provided as a preview when available for these audio-only routes: /audio-inpaint, /audio-extend-suffix, /audio-extend-prefix, /t2a, /a2a, and /v2a. |
| `lora_file` | object {url, content_type, file_name, file_size} | URL for trained LoRA weights in .safetensors format. |
| `config_file` | object {url, content_type, file_name, file_size} | Configuration for inference endpoint setup. |
| `debug_dataset` | object {url, content_type, file_name, file_size} \| null | Debug output available when `debug_dataset` is enabled: an archive that can be downloaded and contains the training data after preprocessing. |

## Related models

- [LTX 2.3 Trainer (V2) - Audio Inpainting](https://doitong.com/developers/models/ltx23-trainer-v2/audio-inpaint.md) (`ltx23-trainer-v2/audio-inpaint`): $0.0023 / step
- [LTX 2.3 Trainer (V2) - Audio+Video Reference IC-LoRA](https://doitong.com/developers/models/ltx23-trainer-v2/ic-lora/av2av.md) (`ltx23-trainer-v2/ic-lora/av2av`): $0.0068 / step
- [LTX 2.3 Trainer (V2) - Audio+Video Reference Transformation](https://doitong.com/developers/models/ltx23-trainer-v2/av2av.md) (`ltx23-trainer-v2/av2av`): $0.0068 / step
- [LTX 2.3 Trainer (V2) - Audio-to-Audio](https://doitong.com/developers/models/ltx23-trainer-v2/a2a.md) (`ltx23-trainer-v2/a2a`): $0.0023 / step
- [LTX 2.3 Trainer (V2) - Audio-to-Audio IC-LoRA](https://doitong.com/developers/models/ltx23-trainer-v2/ic-lora/a2a.md) (`ltx23-trainer-v2/ic-lora/a2a`): $0.0023 / step
- [LTX 2.3 Trainer (V2) - Audio-to-Video](https://doitong.com/developers/models/ltx23-trainer-v2/a2v.md) (`ltx23-trainer-v2/a2v`): $0.006 / step
- [LTX 2.3 Trainer (V2) - Backward Audio Extension](https://doitong.com/developers/models/ltx23-trainer-v2/audio-extend-suffix.md) (`ltx23-trainer-v2/audio-extend-suffix`): $0.0023 / step
- [LTX 2.3 Trainer (V2) - Backward Video Extension](https://doitong.com/developers/models/ltx23-trainer-v2/extend-suffix.md) (`ltx23-trainer-v2/extend-suffix`): $0.0061 / step

## Links

- HTML page: https://doitong.com/developers/models/ltx23-trainer-v2/interpolate
- All models and guides for AI agents: https://doitong.com/developers/llms.txt
- OpenAPI 3.1 with this model's typed request body: https://doitong.com/developers/openapi.json?model=ltx23-trainer-v2/interpolate
- Errors: https://doitong.com/developers/docs/errors
- Model details as JSON: https://api.doitong.com/v1/models/ltx23-trainer-v2/interpolate