# TRELLIS.2 Trainer API

> Trains LoRA adapters for Trellis 2 on a zip of 3D assets, for up to three denoisers

Fine-tunes adapters from a zip at `data_url`, containing raw 3D assets or a preprocessed dataset. `denoisers` selects 1 to 3 targets, each trained as its own adapter. Settings: `resolution` 512 or 1024, `rank` 16, 32 or 64, `learning_rate`, and `training_steps` from 100 to 10000. It returns adapters, any failures and the reusable preprocessed data.

- Model id: `trellis-2-lora-trainer`
- Made by: Microsoft
- Category: Training
- Output: file
- Price per run: from $1.7508 / run (cheapest run $1.751 (1751 API credits))
- Run: `POST https://api.doitong.com/v1/run/trellis-2-lora-trainer`
- Quote: `POST https://api.doitong.com/v1/quote/trellis-2-lora-trainer`
- Docs page: https://doitong.com/developers/models/trellis-2-lora-trainer

## Pricing

Price: **from $1.7508 / run** (the lowest tier; the quote prices your exact input). Price per run: the input files are measured before the run (length, resolution, frame rate). The quote measures the same files and is exactly what the run charges.

The full example input below costs $15.973 = 15973 API credits (1 run × $15.9723).

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/trellis-2-lora-trainer` 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 |
|---|---|---|---|---|---|---|
| `data_url` |  | string | yes |  |  | Dataset archive URL: supply raw 3D assets packaged as a .zip, or use the preprocessed TRELLIS.2 dataset zip returned by data_preprocessing. |
| `denoisers` |  | array of enum (string) | no |  | 1 – 3 items | Selection of 1–3 distinct TRELLIS.2 denoisers from sparse_structure, geometry, and texture for fine-tuning. Parallel training uses a shared preprocessed dataset to produce a separate LoRA adapter for each selected entry. |
| `resolution` |  | enum (integer) | no | `512` | `512`, `1024` | Training resolution must equal the dataset's preprocessing resolution. Geometry and texture training each have separate configurations for 512 and 1024, whereas sparse-structure training shares a configuration across these resolutions. |
| `rank` |  | enum (integer) | no | `32` | `16`, `32`, `64` | Rank of the LoRA adapter; larger values provide greater adapter capacity. |
| `learning_rate` |  | number | no | `0.0001` | 0.000001 – 0.01 | Learning rate used to optimize LoRA. |
| `training_steps` |  | integer | no | `1000` | 100 – 10,000 | Total count of training steps. |

## Example input

Required parameters only:

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

A fuller example:

```json
{
  "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/trellis-2-lora-trainer?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "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/trellis-2-lora-trainer?wait=20", {
  method: "POST",
  headers,
  body: JSON.stringify({
    "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 = {
    "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/trellis-2-lora-trainer", 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": "trellis-2-lora-trainer",
      "input": {
        "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": "trellis-2-lora-trainer",
  "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": 15973,
    "usd": "15.973",
    "charged": 15973,
    "refunded": 0
  },
  "input": {
    "data_url": "https://example.com/your-file.zip"
  },
  "output": {
    "adapters": [
      {
        "denoiser": "sparse_structure",
        "resolution": 1,
        "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
        },
        "rank": 1,
        "training_steps": 1,
        "learning_rate": 1.0
      }
    ],
    "failed": [
      {
        "denoiser": "sparse_structure",
        "error": "string"
      }
    ],
    "preprocessed_data_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
    }
  },
  "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 |
|---|---|---|
| `adapters` | array of object {denoiser, resolution, lora_file, rank, training_steps, learning_rate} | LoRA adapter files mapped to their corresponding trellis-2-lora inference fields. The response includes one adapter for each denoiser whose training succeeded; use its file in that matching field. |
| `failed` | array of object {denoiser, error} | Unsuccessful denoiser training results; only denoisers that failed to finish training appear here, so success across all requested denoisers leaves this empty. |
| `preprocessed_data_file` | object {url, content_type, file_name, file_size} | Preprocessed TRELLIS.2 dataset zip used during training and available for reuse. |

## Related models

- [ERNIE-Image Trainer](https://doitong.com/developers/models/ernie-image-trainer.md) (`ernie-image-trainer`): $0.0012 / step
- [FLUX 2 \[klein\] 9b Base Trainer](https://doitong.com/developers/models/flux-2-klein-9b-base-trainer.md) (`flux-2-klein-9b-base-trainer`): $0.0043 / step
- [Flux 2 Klein 9B Base Trainer](https://doitong.com/developers/models/flux-2-klein-9b-base-trainer/edit.md) (`flux-2-klein-9b-base-trainer/edit`): $0.0139 / step
- [FLUX 2 Trainer](https://doitong.com/developers/models/flux-2-trainer.md) (`flux-2-trainer`): $0.0064 / step
- [FLUX 2 Trainer Edit](https://doitong.com/developers/models/flux-2-trainer/edit.md) (`flux-2-trainer/edit`): $0.0389 / step
- [FLUX 2 Trainer V2](https://doitong.com/developers/models/flux-2-trainer-v2.md) (`flux-2-trainer-v2`): $0.0064 / step
- [FLUX 2 Trainer V2 Edit](https://doitong.com/developers/models/flux-2-trainer-v2/edit.md) (`flux-2-trainer-v2/edit`): $0.0389 / step
- [Flux Kontext Trainer](https://doitong.com/developers/models/flux-kontext-trainer.md) (`flux-kontext-trainer`): from $1.25 / run

## Links

- HTML page: https://doitong.com/developers/models/trellis-2-lora-trainer
- 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=trellis-2-lora-trainer
- Errors: https://doitong.com/developers/docs/errors
- Model details as JSON: https://api.doitong.com/v1/models/trellis-2-lora-trainer