Flux 2 Klein LoRA training

Train a Flux 2 Klein LoRA for use with the Flux 2 image generation recipe. Two size tiers, plus a special edit-training mode for image-editing LoRAs that take control / reference images at inference time.

`modelVariant`	Base	Default price	Use when
`4b` (default)	`FLUX.2-klein-base-4B`	500 Buzz	Cheaper / faster training. Pairs with Klein `4b` inference.
`9b`	`FLUX.2-klein-base-9B`	1000 Buzz	Higher fidelity. Pairs with Klein `9b` inference.

The base checkpoint is fixed by modelVariant; there is no model field on the input. Set isEditTraining: true to train an editing LoRA — the dataset zip layout changes (see Edit training).

Long-running step

Always submit with wait=0. Klein training runs at a fraction of a second per step on a 10-image dataset; a typical 2000-step run lands in single-digit minutes for 4b, longer for 9b.

The request shape

json

{
  "$type": "training",
  "input": {
    "engine":         "ai-toolkit",
    "ecosystem":      "flux2klein",
    "modelVariant":   "4b",         // 4b | 9b
    "isEditTraining": false         // optional, defaults to false
  }
}

Prerequisites

A Civitai orchestration token (Quick start → Prerequisites)
A training-data zip:
- For standard training: a flat zip of training images
- For edit training: a zip with main/, control_1/, control_2/, control_3/ subfolders

Klein 4b (default)

Fastest and cheapest tier. Default for most LoRAs.

http

POST https://orchestration.civitai.com/v2/consumer/workflows?wait=0
Authorization: Bearer <your-token>
Content-Type: application/json

{
  "tags": ["training"],
  "steps": [{
    "$type": "training",
    "priority": "normal",
    "retries": 2,
    "input": {
      "engine": "ai-toolkit",
      "ecosystem": "flux2klein",
      "modelVariant": "4b",
      "steps": 2000,
      "lr": 0.0005,
      "trainTextEncoder": false,
      "lrScheduler": "constant",
      "optimizerType": "adamw8bit",
      "networkDim": 2,
      "networkAlpha": 1,
      "trainingData": {
        "type": "zip",
        "sourceUrl": "urn:air:other:other:civitai-r2:civitai-delivery-worker-prod@training-images/6/2658016TrainingData.1zGG.zip",
        "count": 15
      },
      "samples": {
        "prompts": [
          "fruit, food, no humans, blue eyes, solo, leaf, strawberry, fangs, pokemon (creature)",
          "no humans, pokemon (creature), cup, food, solo, bird, blush, blurry, animal focus",
          "no humans, candle, pokemon (creature), blurry, animal focus, solo, food, bird, standing"
        ]
      }
    }
  }]
}

Set your Civitai API token via the Token button in the navbar to enable Try It.

Request body — edit to customize (e.g. swap the image URL or prompt)

{
  "tags": [
    "training"
  ],
  "steps": [
    {
      "$type": "training",
      "priority": "normal",
      "retries": 2,
      "input": {
        "engine": "ai-toolkit",
        "ecosystem": "flux2klein",
        "modelVariant": "4b",
        "steps": 2000,
        "lr": 0.0005,
        "trainTextEncoder": false,
        "lrScheduler": "constant",
        "optimizerType": "adamw8bit",
        "networkDim": 2,
        "networkAlpha": 1,
        "noiseOffset": 0,
        "flipAugmentation": false,
        "shuffleTokens": false,
        "keepTokens": 0,
        "trainingData": {
          "type": "zip",
          "sourceUrl": "urn:air:other:other:civitai-r2:civitai-delivery-worker-prod@training-images/6/2658016TrainingData.1zGG.zip",
          "count": 15
        },
        "samples": {
          "prompts": [
            "fruit, food, no humans, blue eyes, solo, leaf, strawberry, fangs, pokemon (creature), standing",
            "no humans, pokemon (creature), cup, food, solo, bird, blush, blurry, animal focus, leaf",
            "no humans, candle, pokemon (creature), blurry, animal focus, solo, food, bird, standing"
          ]
        }
      }
    }
  ]
}

Valid JSON

Klein 9b

Same shape, larger base model. Recommended steps: 2000+, networkDim: 32, lr: ~1e-4.

http

POST https://orchestration.civitai.com/v2/consumer/workflows?wait=0
Authorization: Bearer <your-token>
Content-Type: application/json

{
  "tags": ["training"],
  "steps": [{
    "$type": "training",
    "priority": "normal",
    "retries": 2,
    "input": {
      "engine": "ai-toolkit",
      "ecosystem": "flux2klein",
      "modelVariant": "9b",
      "steps": 2000,
      "resolution": 1024,
      "lr": 0.000102,
      "trainTextEncoder": false,
      "lrScheduler": "cosine",
      "optimizerType": "adamw8bit",
      "networkDim": 32,
      "networkAlpha": 32,
      "trainingData": {
        "type": "zip",
        "sourceUrl": "urn:air:other:other:civitai-r2:civitai-delivery-worker-prod@training-images/6/2657604TrainingData.EYBd.zip",
        "count": 1
      },
      "samples": { "prompts": [] }
    }
  }]
}

Set your Civitai API token via the Token button in the navbar to enable Try It.

Request body — edit to customize (e.g. swap the image URL or prompt)

Valid JSON

Edit training

Setting isEditTraining: true produces an editing LoRA — at inference time it takes one or more reference images alongside the prompt and modifies them. The dataset zip layout differs:

main/ — target images (what the LoRA should produce)
control_1/, control_2/, control_3/ — reference / source images that pair with each main/ entry

Filenames inside the subfolders must align across folders. Reading the result you'll get a LoRA that works with Flux 2 Klein → editImage.

http

POST https://orchestration.civitai.com/v2/consumer/workflows?wait=0
Authorization: Bearer <your-token>
Content-Type: application/json

{
  "tags": ["training", "edit"],
  "steps": [{
    "$type": "training",
    "priority": "normal",
    "retries": 2,
    "input": {
      "engine": "ai-toolkit",
      "ecosystem": "flux2klein",
      "modelVariant": "4b",
      "isEditTraining": true,
      "steps": 2000,
      "lr": 0.0001,
      "trainTextEncoder": false,
      "lrScheduler": "cosine",
      "optimizerType": "adamw8bit",
      "networkDim": 32,
      "networkAlpha": 32,
      "trainingData": {
        "type": "zip",
        "sourceUrl": "https://blobs-temp.sfo3.digitaloceanspaces.com/flux2_klein_edit_testdata.zip",
        "count": 3
      },
      "samples": {
        "prompts": [
          "a portrait of a woman standing in a sunlit garden with flowers",
          "a landscape painting of rolling hills at sunset",
          "a painting of a cat sitting on a windowsill looking outside at a rainy day"
        ],
        "sourceImages": [
          "https://blobs-temp.sfo3.digitaloceanspaces.com/sample-edit-source-1.jpg",
          "https://blobs-temp.sfo3.digitaloceanspaces.com/sample-edit-source-2.jpg"
        ]
      }
    }
  }]
}

Set your Civitai API token via the Token button in the navbar to enable Try It.

Request body — edit to customize (e.g. swap the image URL or prompt)

{
  "tags": [
    "training",
    "edit"
  ],
  "steps": [
    {
      "$type": "training",
      "priority": "normal",
      "retries": 2,
      "input": {
        "engine": "ai-toolkit",
        "ecosystem": "flux2klein",
        "modelVariant": "4b",
        "isEditTraining": true,
        "steps": 2000,
        "lr": 0.0001,
        "trainTextEncoder": false,
        "lrScheduler": "cosine",
        "optimizerType": "adamw8bit",
        "networkDim": 32,
        "networkAlpha": 32,
        "noiseOffset": 0,
        "trainingData": {
          "type": "zip",
          "sourceUrl": "https://blobs-temp.sfo3.digitaloceanspaces.com/flux2_klein_edit_testdata.zip",
          "count": 3
        },
        "samples": {
          "prompts": [
            "a portrait of a woman standing in a sunlit garden with flowers",
            "a landscape painting of rolling hills at sunset",
            "a painting of a cat sitting on a windowsill looking outside at a rainy day"
          ],
          "sourceImages": [
            "https://blobs-temp.sfo3.digitaloceanspaces.com/sample-edit-source-1.jpg",
            "https://blobs-temp.sfo3.digitaloceanspaces.com/sample-edit-source-2.jpg"
          ]
        }
      }
    }
  ]
}

Valid JSON

samples.sourceImages is required for edit training when you want preview samples — the listed URLs become the reference images for the per-checkpoint sample renders.

Common parameters

Defaults shown are the post-ApplyDefaults values for Klein.

Field	Required	Default	Notes
`engine`	✅	—	Always `ai-toolkit`.
`ecosystem`	✅	—	Always `flux2klein` for this page.
`modelVariant`	✅	—	`4b` or `9b`.
`isEditTraining`		`false`	When `true`, dataset zip must contain `main/` + `control_*/` subfolders.
`steps`		`2000`	`1`–`10000`. Total training steps. Primary driver of training length and pricing.
`epochs`		`10`	`1`–`20`. Number of saved checkpoints delivered, each separately downloadable. Each adds 10 Buzz of storage.
`batchSize`		`1`	Defaults to 1. For `4b`, raise it up to the ecosystem maximum (2) to train faster at the cost of more GPU memory; a larger batch needs fewer steps. For `9b` it is fixed at 1. Values above the max are clamped.
`continueFrom`		(none)	A previously-trained `urn:air:flux2:lora:...` AIR to resume from (see Continue training). Must be a Klein (flux2) LoRA.
`lr`		`0.0001`	Klein is sensitive to high LRs — keep in `1e-4`–`5e-4`.
`trainTextEncoder`		`false`	Klein uses Qwen-3 as its text encoder; AI Toolkit does not train it. Leave `false`.
`lrScheduler`		`cosine`	`constant`, `constant_with_warmup`, `cosine`, `linear`, `step`.
`optimizerType`		`adamw8bit`	See SDXL/SD1 page for full enum.
`networkDim`		`32`	`1`–`256`. Klein LoRAs are typically `16`–`32`.
`networkAlpha`		matches `networkDim`	`1`–`256`.
`noiseOffset`		`0`	`0`–`1`.
`flipAugmentation`		`false`	Random horizontal flips.
`shuffleTokens` / `keepTokens`		`false` / `0`	Caption-tag shuffling.
`triggerWord`		(none)	Activation token.
`trainingData.{type, sourceUrl, count}`	✅	—	`type: "zip"`. For edit training, `count` should equal the number of `main/` entries.
`samples.prompts[]`		`[]`	Preview prompts rendered at each saved checkpoint.
`samples.negativePrompt`		(none)	—
`samples.cfgScale`		(ecosystem default)	Overrides the CFG / guidance scale used when rendering the preview samples.
`samples.strength`		`1.0`	Trained-LoRA weight applied in the preview samples.
`samples.sourceImages[]`		`[]`	Edit-training only — reference images for sample renders.

Continue training / train further

To resume from a Klein LoRA you already trained instead of starting from the base checkpoint, set continueFrom to that LoRA's AIR. The new run starts from those weights and the new epochs build on top:

json

{
  "$type": "training",
  "input": {
    "engine": "ai-toolkit",
    "ecosystem": "flux2klein",
    "modelVariant": "4b",
    "continueFrom": "urn:air:flux2:lora:civitai:<id>@<version>",
    "steps": 1000
  }
}

continueFrom must point at a LoRA of the same ecosystem (a Klein / flux2 LoRA) as the model being trained — a mismatched ecosystem is rejected.

Reading the result

Same envelope as the other training recipes — see SDXL/SD1 → Reading the result. Each saved checkpoint yields a Klein LoRA .safetensors blob plus any sample images.

To use the trained LoRA, register it on Civitai (or reference its blob URN directly) and pass it under loras in a Flux 2 Klein generation request:

json

{
  "$type": "imageGen",
  "input": {
    "engine": "flux2",
    "model": "klein",
    "operation": "createImage",
    "modelVersion": "4b",
    "prompt": "your prompt",
    "loras": { "urn:air:flux2:lora:civitai:<id>@<version>": 1.0 }
  }
}

Runtime

Per-step wall time, default settings on a 10-image dataset:

Variant	Per-step	Typical full run
`4b`	~0.1–0.3 s	3–10 min for 2000 steps
`9b`	~0.3–0.6 s	10–30 min for 2000 steps
`4b` + `isEditTraining`	~0.2–0.5 s	5–15 min for 2000 steps

Always use wait=0.

Cost

Training is billed per step plus a flat per-epoch storage surcharge, with a price floor:

price = steps × costPerStep + epochs × 10        (rounded)
costPerStep = 0.20 (4b), 0.45 (9b)
floor: never less than 80% of the default-configuration price

epochs is the number of saved checkpoints delivered (default 10, range 1–20); each adds 10 Buzz of storage. The default run is 2000 steps / 10 epochs:

4b: 2000 × 0.20 + 10 × 10 = 400 + 100 = 500 Buzz (floor 400)
9b: 2000 × 0.45 + 10 × 10 = 900 + 100 = 1000 Buzz (floor 800)

The floor is 80% of the default price — lowering steps or epochs can save at most 20%.

Configuration	Buzz
Klein `4b`, default (`steps: 2000`, `epochs: 10`)	500 + samples
Klein `4b`, `steps: 1000`, `epochs: 10`	400 + samples (floor)
Klein `9b`, default (`steps: 2000`, `epochs: 10`)	1000 + samples
Klein `9b`, `steps: 1000`, `epochs: 10`	800 + samples (floor)

Sample-prompt rendering is billed separately at Klein image-generation rates (~8 Buzz per sample for 4b, ~16 for 9b). Run with whatif=true to see exact charges.

Troubleshooting

Symptom	Likely cause	Fix
`400` with "modelVariant required"	Missing `modelVariant`	Set to `"4b"` or `"9b"`.
`400` with "isEditTraining: true requires control folders"	Edit-training zip missing `control_*/` subfolders	Repackage the zip with `main/`, `control_1/`, `control_2/`, `control_3/`. Filenames must align across folders.
Step `failed` mentioning "training data validation"	Edit-training zip filenames don't match across `main/` and `control_*/`	Ensure the same basenames appear in `main/` and at least one `control_*/` folder.
Trained LoRA underbaked	Too few steps / too low `lr`	Raise `steps` to 1500–2500; keep `lr` ≤ `5e-4`.
Trained LoRA overcooked / broken samples	`lr` too high	Drop `lr` to `1e-4`–`2e-4`.
Step `failed`, `moderationStatus: "Rejected"`	Dataset failed content moderation	Replace flagged images.

Flux 1 LoRA training — open-weights Flux LoRAs (Dev / Schnell)
SDXL & SD1 LoRA training — cheaper SD-family ecosystems
Flux 2 image generation — use a trained Klein LoRA via loras: { ... } on model: "klein"
Results & webhooks — handling long-running training jobs
SubmitWorkflow / GetWorkflow
Endpoint OpenAPI spec

Flux 2 Klein LoRA training ​

The request shape ​

Prerequisites ​

Klein 4b (default) ​

Klein 9b ​

Edit training ​

Common parameters ​

Continue training / train further ​

Reading the result ​

Runtime ​

Cost ​

Troubleshooting ​

Related ​

Flux 2 Klein LoRA training

The request shape

Prerequisites

Klein 4b (default)

Klein 9b

Edit training

Common parameters

Continue training / train further

Reading the result

Runtime

Cost

Troubleshooting

Related