# Ideogram v4.0q LoRA trainer

Needs images data url. Served by fal.ai.

Ideogram V4.0q LoRA Trainer, served by fal.ai. The OpenAPI schema for the ideogram/v4/trainer queue. Price: The cost of training depends on the number of steps. The formula is: 0.00675 * steps. With 1000 steps, your request will cost $6.75. Queued execution; final USD charge comes from the unique billing event matching this exact endpoint and request_id. A missing receipt retains the reservation; no guessed final fee. Asynchronous: returns a request_id; poll fal/status and fetch fal/result with model "ideogram/v4" and that request_id. Billed once, when the result is ready; a generation that fails is refunded.

- Tool ID: `fal/ideogram-v4-trainer`
- Name in the catalog: Ideogram V4.0q LoRA Trainer
- Provider: [fal.ai](/providers/fal.md)
- Capability: [Ideogram](/tools/image-generation/ideogram.md)
- Price: $0.00675 per step
- Authentication: an authorised Agentik workspace is required for execution, including zero-price tools.

## Pricing contract

The provider’s price, with no markup. Billed for measured usage.


```json
{
  "model": "per_unit",
  "amountUsd": 0.00675,
  "unitsField": "steps",
  "unitsDefault": 1000,
  "maxUnits": 40000,
  "fractional": true
}
```

Pricing source: [https://fal.ai/models/ideogram/v4/trainer](https://fal.ai/models/ideogram/v4/trainer). The cost of training depends on the number of steps. The formula is: 0.00675 * steps. With 1000 steps, your request will cost $6.75. Queued execution; final USD charge comes from the unique billing event matching this exact endpoint and request_id. A missing receipt retains the reservation; no guessed final fee..

Price verification date: 2026-10-07.

## Complete input schema


```json
{
  "type": "object",
  "properties": {
    "learning_rate": {
      "title": "Learning Rate",
      "type": "number",
      "minimum": 0.000001,
      "maximum": 0.01,
      "default": 0.0001
    },
    "output_lora_format": {
      "title": "Output Lora Format",
      "type": "string",
      "description": "Dictates the naming scheme for the output weights",
      "enum": [
        "fal",
        "comfy"
      ],
      "default": "fal"
    },
    "steps": {
      "title": "Steps",
      "maximum": 40000,
      "minimum": 100,
      "type": "integer",
      "description": "Number of LoRA training steps.",
      "default": 1000
    },
    "images_data_url": {
      "title": "Images Data Url",
      "type": "string",
      "description": "URL to a zip archive with training images and captions. The trainer uses the target-only dataset prepared by prepare_dataset_dir. Captions are read from same-stem .txt files."
    },
    "default_caption": {
      "title": "Default Caption",
      "description": "Caption to use when an image has no caption file. If force_default_caption is true, this caption replaces all caption files.",
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ]
    },
    "resolution": {
      "title": "Resolution",
      "type": "string",
      "description": "Training image size. Use auto to pick the largest common no-upscale crop from the dataset. Images are center-cropped without upscaling. Available presets: square 1024x1024, landscape 1536x1024, portrait 1024x1536, widescreen 1920x1088, ultrawide 2048x768, phone_wallpaper 1024x1792, social_banner 1584x400. You can also pass a custom WIDTHxHEIGHT string when both values are divisible by 16.",
      "default": "auto"
    }
  },
  "required": [
    "images_data_url"
  ],
  "additionalProperties": false
}
```

## Published example

This is the provider tool’s published example. Replace example values with the user’s requested inputs, respecting the schema; do not execute an example merely to test a connection.


```json
{
  "tool_id": "fal/ideogram-v4-trainer",
  "input": {
    "images_data_url": ""
  }
}
```

## Execute and retrieve

Inspect again before execution, explain the price and obtain the required authority. Use MCP run_tool or POST /v1/run with this tool ID, valid input and an idempotency key. Save the returned run ID and follow the actual state. If Agentik answers that the tool is temporarily unavailable, do not retry it blindly; use discovery to propose an alternative.

This tool uses deferred settlement. Follow its returned task/run instructions; preserve the existing reservation until the result is reconciled.

A partial or uncertain result can still have consumed provider resources. Consult [billing](/docs/billing.md), [idempotency](/docs/idempotency.md) and [asynchronous results](/docs/async.md).

[Provider documentation](https://fal.ai/models/ideogram/v4/trainer)

[REST inspection](https://api.agentik.cc/v1/tools/fal/ideogram-v4-trainer) · [OpenAPI](/openapi.json) · [Connection guide](/connect.md)

---

[HTML page](/providers/fal/ideogram-v4-trainer) · [Agent index](/llms.txt)