# FLUX.2 [klein] 9B Base LoRA > Image-to-image editing with LoRA support for FLUX.2 [klein] 9B Base from Black Forest Labs. Specialized style transfer and domain-specific modifications. ## Overview - **Endpoint**: `https://fal.run/fal-ai/flux-2/klein/9b/base/edit/lora` - **Model ID**: `fal-ai/flux-2/klein/9b/base/edit/lora` - **Category**: image-to-image - **Kind**: inference ## Pricing Requests cost **$0.02** per megapixel of input and output. Input images will be resized to 1MP. For example, a 1024x1024 generation with 512*512 size image input will cost **0.04** (1 MP input + 1 MP output, each costing **0.02**). For more details, see [fal.ai pricing](https://fal.ai/pricing). ## API Information This model can be used via our HTTP API or more conveniently via our client libraries. See the input and output schema below, as well as the usage examples. ### Input Schema The API accepts the following input parameters: - **`prompt`** (`string`, _required_): The prompt to edit the image. - Examples: "Change his clothes to casual suit and tie" - **`negative_prompt`** (`string`, _optional_): Negative prompt for classifier-free guidance. Describes what to avoid in the image. Default value: `""` - Default: `""` - **`guidance_scale`** (`float`, _optional_): Guidance scale for classifier-free guidance. Default value: `5` - Default: `5` - Range: `0` to `20` - **`seed`** (`integer`, _optional_): The seed to use for the generation. If not provided, a random seed will be used. - **`num_inference_steps`** (`integer`, _optional_): The number of inference steps to perform. Default value: `28` - Default: `28` - Range: `4` to `50` - **`image_size`** (`ImageSize | Enum`, _optional_): The size of the generated image. If not provided, uses the input image size. - One of: ImageSize | Enum - Examples: {"height":1152,"width":2016} - **`num_images`** (`integer`, _optional_): The number of images to generate. Default value: `1` - Default: `1` - Range: `1` to `4` - **`acceleration`** (`AccelerationEnum`, _optional_): The acceleration level to use for image generation. Default value: `"regular"` - Default: `"regular"` - Options: `"none"`, `"regular"`, `"high"` - **`sync_mode`** (`boolean`, _optional_): If `True`, the media will be returned as a data URI. Output is not stored when this is True. - Default: `false` - **`enable_safety_checker`** (`boolean`, _optional_): If set to true, the safety checker will be enabled. Default value: `true` - Default: `true` - **`output_format`** (`OutputFormatEnum`, _optional_): The format of the generated image. Default value: `"png"` - Default: `"png"` - Options: `"jpeg"`, `"png"`, `"webp"` - **`enable_cns`** (`boolean`, _optional_): Enable Colored Noise Sampling (CNS), a stochastic SDE sampler. Off by default; when enabled the cns_* parameters below take effect. - Default: `false` - **`cns_s_churn`** (`float`, _optional_): How much stochastic (colored) noise is injected after each step. 0 = CNS off, i.e. pure deterministic flow-matching (identical to normal klein); higher = more stochasticity. Turn this first; the other CNS parameters only matter when s_churn > 0. Default value: `0.5` - Default: `0.5` - Range: `0` to `2` - **`cns_gamma_source`** (`CNSGammaSourceEnum`, _optional_): Where the per-frequency progress signal gamma(f, t) comes from. 'approximation' is self-contained but uniform across bands, so the frequency shaping comes entirely from the alpha-tilt parameters. 'matrix' uses the precomputed per-band gamma for true spectral-bias-aware coloring (more noise to genuinely under-resolved bands). Default value: `"matrix"` - Default: `"matrix"` - Options: `"matrix"`, `"approximation"` - **`cns_power_gamma`** (`float`, _optional_): Exponent on the residual (1 - gamma). Higher concentrates noise more aggressively on unresolved bands; lower is gentler and flatter. No effect under 'approximation', where gamma is uniform. Default value: `0.75` - Default: `0.75` - Range: `0.1` to `3` - **`cns_gamma_divider`** (`float`, _optional_): Divides gamma first, weakening the coloring (pushing it toward uniform white noise). Paper uses 1.73 for distilled/unguided and 25.0 for CFG/base, so raise this for the /base (CFG) endpoints. Default value: `1.73` - Default: `1.73` - Range: `0.1` to `50` - **`cns_alpha_tilt_start`** (`float`, _optional_): Frequency tilt at the start of sampling, interpolated toward the end value over steps. Positive boosts high frequencies (fine texture/detail); negative boosts low frequencies (structure). Default value: `0.15` - Default: `0.15` - Range: `-2` to `2` - **`cns_alpha_tilt_end`** (`float`, _optional_): Frequency tilt at the end of sampling. Positive boosts high frequencies (fine texture/detail); negative boosts low frequencies (structure). The default favors low-frequency structure at the end. Default value: `-0.5` - Default: `-0.5` - Range: `-2` to `2` - **`cns_alpha_use_fnorm`** (`boolean`, _optional_): Use a smooth, frequency-position-weighted tilt (exp(alpha * f)) instead of a flat (1 + alpha) multiplier. Default value: `true` - Default: `true` - **`cns_alpha_exp_interp`** (`boolean`, _optional_): Use an exponential (vs linear) curve for the start-to-end tilt transition across sampling steps. Default value: `true` - Default: `true` - **`cns_alpha_exp_sharpness`** (`float`, _optional_): How sharp the exponential tilt-interpolation curve is. Only used when exponential interpolation is enabled. Default value: `0.75` - Default: `0.75` - Range: `0.1` to `10` - **`cns_num_freq_bins`** (`integer`, _optional_): How many radial frequency bands the spectrum is split into. More bands = finer-grained coloring. Default value: `32` - Default: `32` - Range: `8` to `128` - **`cns_energy_scale`** (`float`, _optional_): Global multiplier on the injected colored noise after variance normalisation. <1 slightly damps the kick, >1 amplifies it (paper: 0.98 unguided, 0.998 guided). Default value: `0.98` - Default: `0.98` - Range: `0.5` to `1.5` - **`image_urls`** (`list`, _required_): The URLs of the images for editing. A maximum of 4 images are allowed. - Array of string - Examples: ["https://storage.googleapis.com/falserverless/example_inputs/flux2_dev_edit_input.png"] - **`loras`** (`list`, _optional_): List of LoRA weights to apply (maximum 3). - Default: `[]` - Array of fal-ai_flux-2-klein_LoRAInput **Required Parameters Example**: ```json { "prompt": "Change his clothes to casual suit and tie", "image_urls": [ "https://storage.googleapis.com/falserverless/example_inputs/flux2_dev_edit_input.png" ] } ``` **Full Example**: ```json { "prompt": "Change his clothes to casual suit and tie", "guidance_scale": 5, "num_inference_steps": 28, "image_size": { "height": 1152, "width": 2016 }, "num_images": 1, "acceleration": "regular", "enable_safety_checker": true, "output_format": "png", "cns_s_churn": 0.5, "cns_gamma_source": "matrix", "cns_power_gamma": 0.75, "cns_gamma_divider": 1.73, "cns_alpha_tilt_start": 0.15, "cns_alpha_tilt_end": -0.5, "cns_alpha_use_fnorm": true, "cns_alpha_exp_interp": true, "cns_alpha_exp_sharpness": 0.75, "cns_num_freq_bins": 32, "cns_energy_scale": 0.98, "image_urls": [ "https://storage.googleapis.com/falserverless/example_inputs/flux2_dev_edit_input.png" ], "loras": [] } ``` ### Output Schema The API returns the following output format: - **`images`** (`list`, _required_): The generated images - Array of ImageFile - **`timings`** (`Timings`, _required_) - **`seed`** (`integer`, _required_): Seed of the generated Image. It will be the same value of the one passed in the input or the randomly generated that was used in case none was passed. - **`has_nsfw_concepts`** (`list`, _required_): Whether the generated images contain NSFW concepts. - Array of boolean - **`prompt`** (`string`, _required_): The prompt used for generating the image. **Example Response**: ```json { "images": [ { "url": "", "content_type": "image/png", "file_name": "z9RV14K95DvU.png", "file_size": 4404019 } ], "prompt": "" } ``` ## Usage Examples ### cURL ```bash curl --request POST \ --url https://fal.run/fal-ai/flux-2/klein/9b/base/edit/lora \ --header "Authorization: Key $FAL_KEY" \ --header "Content-Type: application/json" \ --data '{ "prompt": "Change his clothes to casual suit and tie", "image_urls": [ "https://storage.googleapis.com/falserverless/example_inputs/flux2_dev_edit_input.png" ] }' ``` ### Python Ensure you have the Python client installed: ```bash pip install fal-client ``` Then use the API client to make requests: ```python import fal_client def on_queue_update(update): if isinstance(update, fal_client.InProgress): for log in update.logs: print(log["message"]) result = fal_client.subscribe( "fal-ai/flux-2/klein/9b/base/edit/lora", arguments={ "prompt": "Change his clothes to casual suit and tie", "image_urls": ["https://storage.googleapis.com/falserverless/example_inputs/flux2_dev_edit_input.png"] }, with_logs=True, on_queue_update=on_queue_update, ) print(result) ``` ### JavaScript Ensure you have the JavaScript client installed: ```bash npm install --save @fal-ai/client ``` Then use the API client to make requests: ```javascript import { fal } from "@fal-ai/client"; const result = await fal.subscribe("fal-ai/flux-2/klein/9b/base/edit/lora", { input: { prompt: "Change his clothes to casual suit and tie", image_urls: ["https://storage.googleapis.com/falserverless/example_inputs/flux2_dev_edit_input.png"] }, logs: true, onQueueUpdate: (update) => { if (update.status === "IN_PROGRESS") { update.logs.map((log) => log.message).forEach(console.log); } }, }); console.log(result.data); console.log(result.requestId); ``` ## Additional Resources ### Documentation - [Model Playground](https://fal.ai/models/fal-ai/flux-2/klein/9b/base/edit/lora) - [API Documentation](https://fal.ai/models/fal-ai/flux-2/klein/9b/base/edit/lora/api) - [OpenAPI Schema](https://fal.ai/api/openapi/queue/openapi.json?endpoint_id=fal-ai/flux-2/klein/9b/base/edit/lora) ### fal.ai Platform - [Platform Documentation](https://docs.fal.ai) - [Python Client](https://docs.fal.ai/clients/python) - [JavaScript Client](https://docs.fal.ai/clients/javascript)