Image
FLUX.2 Image Generator
Type a prompt and download a picture. Free text to image on devices you connect.
FLUX.2 klein by Black Forest Labs
FLUX.2 klein by Black Forest Labs; Qwen3 by Alibaba Cloud
Model licenses
FLUX.2 klein 4B fifteen-stage transformer · Apache-2.0
Retain full Apache-2.0 terms and copyright notices, mark modifications; patent termination and trademark limitations apply.
FLUX.2 klein VAE and shared pipeline/tokenizer derivatives · Apache-2.0
Retain complete Apache-2.0 licence, upstream copyright and NOTICE texts; prominently identify modified files. Patent termination and trademark limitations apply.
Qwen3-4B text encoder distributed by FLUX.2 klein · Apache-2.0
API reference
Submit via REST
Use your personal API key (msk_…) to create jobs for this workload from your own tools. Jobs started this way run on the same devices you connected on Your Devices — at least one has to be online.
https://cp.meshlabs.cloud/v1/organizations/jobsAuthenticate with X-Organization-Api-Key: msk_your_personal_api_key. The header name matches org keys; personal keys use the same endpoint.
POST https://cp.meshlabs.cloud/v1/organizations/job-inputs/presign with the same API key, upload your file to the returned upload_url, then pass download_url in job_spec (for example as input_image_url).Request body
{
"workload_type": "flux-2-klein-4b-t2i",
"job_spec": {
"prompt": "A photorealistic lighthouse at dusk",
"width": 512,
"height": 512,
"strength": 0.8,
"numSteps": 4,
"compression_method": "fp16",
"compression_error_threshold": 0.05,
"compression_hq8_block": 128
}
}job_spec fields
| Field | Required | Default | Description |
|---|---|---|---|
prompt | Yes | — | Text prompt for image generation |
input_image_url | No | — | Optional input image for image editing (img2img). When set, the image is VAE-encoded and mixed with noise by 'strength'; the prompt describes the desired edit. |
width | No | 512 | Output image width in pixels |
height | No | 512 | Output image height in pixels |
seed | No | — | Random seed for reproducibility |
strength | No | 0.8 | img2img noise fraction: 0.3 = light touch-up, 0.8 = heavy edit, 1.0 = ignore the input image. Only used with input_image_url. |
numSteps | No | 4 | Number of denoising steps (FLUX.2-klein default: 4) |
num_steps | No | — | Snake-case alias of numSteps. Omit it and numSteps is what runs. |
compression_method | No | fp16 | Codec for DiT inter-shard activation handoff. 'off' sends raw Float32. hq8/hq4 = Hadamard-quantized; stp = low-rank; fp16 = half precision; auto8 = best-of router. Reported via codec_ms / transport_savings_pct job metrics. |
compression_error_threshold | No | 0.05 | Relative Frobenius error budget for lossy codecs (0.01 strict … 0.05 lenient). |
compression_hq8_block | No | 128 | Hadamard/quant block size for HQ8/HQ4 (power of 2). Smaller = lower error. |
cURL
curl -X POST https://cp.meshlabs.cloud/v1/organizations/jobs \
-H "X-Organization-Api-Key: msk_your_personal_api_key" \
-H "Content-Type: application/json" \
-d '{
"workload_type": "flux-2-klein-4b-t2i",
"job_spec": {
"prompt": "A photorealistic lighthouse at dusk",
"width": 512,
"height": 512,
"strength": 0.8,
"numSteps": 4,
"compression_method": "fp16",
"compression_error_threshold": 0.05,
"compression_hq8_block": 128
}
}'https://cp.meshlabs.cloud/v1/organizations/jobs/{job_id}Poll with the same API key until status is terminal (COMPLETED, FAILED, CANCELED, or MODERATION_BLOCKED). When content moderation is enabled, jobs may start in MODERATION_PENDING before entering the compute queue. Completed jobs include presigned download_url values on each output.
Create response
{
"id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
"workload_type": "flux-2-klein-4b-t2i",
"status": "PENDING",
"job_spec_json": {
"prompt": "A photorealistic lighthouse at dusk",
"width": 512,
"height": 512,
"strength": 0.8,
"numSteps": 4,
"compression_method": "fp16",
"compression_error_threshold": 0.05,
"compression_hq8_block": 128
},
"work_units": [],
"outputs": [],
"created_at": "2026-06-08T12:00:00Z",
"updated_at": "2026-06-08T12:00:00Z"
}Completed response
{
"id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
"workload_type": "flux-2-klein-4b-t2i",
"status": "COMPLETED",
"job_spec_json": {
"prompt": "A photorealistic lighthouse at dusk",
"width": 512,
"height": 512,
"strength": 0.8,
"numSteps": 4,
"compression_method": "fp16",
"compression_error_threshold": 0.05,
"compression_hq8_block": 128
},
"work_units": [
{
"id": "…",
"index": 0,
"status": "COMPLETED"
}
],
"outputs": [
{
"id": "…",
"output_type": "<output type>",
"filename": "<output filename>",
"size_bytes": 524288,
"download_url": "<presigned output download URL>"
}
],
"created_at": "2026-06-08T12:00:00Z",
"updated_at": "<completion timestamp>"
}Type a prompt and download a picture. FLUX.2 Klein is a free online text-to-image model on devices you connect. The first run downloads the model; later pictures start faster. This is a generator from a blank prompt, not an instruction editor.