Skip to content
WorkflowListedExternal
Open source

Flux.2 Klein Photorealism Dual Stage Workflow — v1.0 - T2I

This is a 2-stage workflow which means it starts with the base model at high CFG for strong prompt adherence, then switches to the distilled model (at a specific sigma boundary, around 0.95 works well with 5-6 steps)...

Sourcecivitai.com/models/2432328
FamilyFLUXVersionRelease 2734892 — Flux.2_Klein_Photorealism_Dual_Model_Workflow.jsonEstimated VRAMNot yet characterizedBgrade

Runtime profile

Access
External workflow
Status
Listed
Version
Flux.2_Klein_Photorealism_Dual_Model_Workflow.json
Source
Civitai
Updated
Sep 30, 2026

Source description

This is a 2-stage workflow which means it starts with the base model at high CFG for strong prompt adherence, then switches to the distilled model (at a specific sigma boundary, around 0.95 works well with 5-6 steps) for faster processing (CFG 1) and fine details, similar to WAN's high/low workflows.

Uses this lora for the 2nd (distilled) stage: https://civitai.com/models/2401004?modelVersionId=2699601

What you need

Model files
17.2 GiB known subtotal

Remaining hardware or download-size estimates are not available yet.

Requirements

Models

4

Node packs

2

AI-generated commentary

AI-generated explanation based on source and configuration details. Suggestions are clearly labeled.

Purpose

  • This two-stage workflow is for prompt-guided photorealistic image generation: it takes a prompt, starts with Flux.2 Klein 9B-base at high CFG, switches to a distilled model at CFG 1, and is intended to produce images with fine details.

Flow

  • Stage one uses the base model at high CFG for strong prompt adherence.

  • At a sigma boundary around 0.95, stage two switches to the distilled model at CFG 1 for fine details.

Inputs and outputs

  • Input: a prompt.

  • Output: a photorealistic image with fine details.

Setup

  • Published by ShaftOverflow and based on Flux.2 Klein 9B-base.

  • No license is listed for this workflow.

Dependencies

  • It requires the node packs comfyui-easy-use and rgthree-comfy.

  • The listed model files are flux-2-klein-9b.safetensors, flux-2-klein-base-9b.safetensors, and flux2-vae.safetensors.

  • The listed text encoder file is qwen_3_8b.safetensors.

Configuration

  • Use high CFG for the base stage and CFG 1 for the distilled stage.

  • A suggested starting point is a sigma boundary around 0.95 with 5–6 steps.

    Suggestion · not verified

  • Use this LoRA for the second, distilled stage: https://civitai.com/models/2401004?modelVersionId=2699601

Cautions

  • Before use, verify that the model filenames match your intended setup.

    Suggestion · not verified

Does this need to be edited?

Sign in to send an edit request.

Sources

1 source
CivitaiSourceVersion 2734892
civitai.com/models/2432328

Verified Sep 30

Source excerpts

3 excerpts
NoteSep 30civitai.com/models/2432328

Source context: 411 downloads · Type Workflows · Base model Flux.2 Klein 9B-base

NoteSep 28civitai.com/models/2432328

Source context: 408 downloads · Type Workflows · Base model Flux.2 Klein 9B-base

NoteJul 18civitai.com/models/2432328

Source context: 358 downloads · Type Workflows · Base model Flux.2 Klein 9B-base