Edit a picture in plain words or make a new one from scratch. For an edit, you can give P-H3 up to seven more pictures and tell it what to take from each one.P-H3 makes its best pictures at 8 MP. Smaller sizes can give blurry faces and artifacts. The Size pill sets edits and new pictures to 4 MP or 8 MP: pick 8 MP if your card can take it.<h3 id="how-to-use-it">How to use it</h3><ol><li>Drop the workflow into ComfyUI. Load the picture you want to change and write your edit.</li><li>For a reference, connect an image loader to picture 2 on Write the prompt. Add more in order through picture 8. Name the picture in your request, for example “Give the person in picture 1 the jacket from picture 2.”</li><li>Press Run. Drag the comparison slider to check what changed. Auto prompt shows the expanded request in Show Text.</li><li>To generate from words, turn Edit a picture off and choose a shape in Latent Size. The workflow scales it to the Size you picked. Turn Upscale the result on if you want SeedVR2 to double the width and height.</li></ol><h3 id="what-you-need">What you need</h3>Update ComfyUI and <a target="_blank" rel="ugc" href="https://github.com/ausboss/ComfyUI-AusBoss">AusBoss nodes</a> first. This version ran on ComfyUI 0.44.0 with AusBoss nodes 2.20.5. Pick one matching model and text encoder pair. The workflow opens on Low VRAM.<ul><li>Low VRAM · <a target="_blank" rel="ugc" href="https://huggingface.co/PrunaAI/P-H3-Image-Comfy/resolve/main/diffusion_models/p-h3-image_w4a8.safetensors">p-h3-image_w4a8.safetensors</a> → <code>models/diffusion_models</code> · 13.4 GB</li><li>Low VRAM · <a target="_blank" rel="ugc" href="https://huggingface.co/PrunaAI/P-H3-Image-Comfy/resolve/main/text_encoders/p-h3-image_text_encoder_w4a8.safetensors">p-h3-image_text_encoder_w4a8.safetensors</a> → <code>models/text_encoders</code> · 3.6 GB</li><li>High VRAM · <a target="_blank" rel="ugc" href="https://huggingface.co/PrunaAI/P-H3-Image-Comfy/resolve/main/diffusion_models/p-h3-image_int8_convrot.safetensors">p-h3-image_int8_convrot.safetensors</a> → <code>models/diffusion_models</code> · 21.2 GB</li><li>High VRAM · <a target="_blank" rel="ugc" href="https://huggingface.co/PrunaAI/P-H3-Image-Comfy/resolve/main/text_encoders/p-h3-image_text_encoder_int8_convrot.safetensors">p-h3-image_text_encoder_int8_convrot.safetensors</a> → <code>models/text_encoders</code> · 5.1 GB</li><li>Both · <a target="_blank" rel="ugc" href="https://huggingface.co/PrunaAI/P-H3-Image-Comfy/resolve/main/vae/p-h3-image_vae_int8_convrot.safetensors">p-h3-image_vae_int8_convrot.safetensors</a> → <code>models/vae</code> · 3.1 GB</li><li>Low VRAM · Auto prompt · <a target="_blank" rel="ugc" href="https://huggingface.co/PrunaAI/P-H3-Image-Comfy/resolve/main/text_encoders/p-h3-image_prompt_upsampler_w4a8.safetensors">p-h3-image_prompt_upsampler_w4a8.safetensors</a> → <code>models/text_encoders</code> · 6.7 GB</li><li>High VRAM · Auto prompt · <a target="_blank" rel="ugc" href="https://huggingface.co/PrunaAI/P-H3-Image-Comfy/resolve/main/text_encoders/p-h3-image_prompt_upsampler_bf16.safetensors">p-h3-image_prompt_upsampler_bf16.safetensors</a> → <code>models/text_encoders</code> · 19.3 GB</li><li>2× upscale · <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/SeedVR2/resolve/main/diffusion_models/seedvr2_3b_int8_convrot.safetensors">seedvr2_3b_int8_convrot.safetensors</a> → <code>models/diffusion_models</code> · 3.46 GB</li><li>2× upscale · <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/SeedVR2/resolve/main/vae/seedvr2_ema_vae_fp16.safetensors">seedvr2_ema_vae_fp16.safetensors</a> → <code>models/vae</code> · 501 MB</li></ul><h3 id="tips">Tips</h3><ul><li>With Auto prompt off, say what should stay the same as well as what you want to change.</li><li>Check faces, lighting and the background after a multi-picture edit. The model can change more than you asked for.</li><li>Size: 4 MP | 8 MP sets the size of an edit and of a new picture. 4 MP is the default and uses less memory. 8 MP can improve faces and fine detail, with more memory and time. P-H3 makes its best new pictures at 8 MP.</li><li>Higher resolution does not lock identity. Check the face against the original after an edit.</li></ul>Powered by MiniMax H3. P-H3-Image is by Pruna AI and uses the <a target="_blank" rel="ugc" href="https://huggingface.co/PrunaAI/P-H3-Image-Comfy/blob/main/LICENSE">MiniMax H3 Community License</a>. Optional upscaling uses <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/SeedVR2">SeedVR2</a> under Apache 2.0.v1.1 (2026-11-07) · Added the Size choice, 4 MP or 8 MP, for edits and new pictures. Low VRAM now uses the smaller prompt writer.v1.0 (2026-11-06) · First release.<a target="_blank" rel="ugc" href="https://github.com/ausboss/ComfyUI-AusBoss/issues">Report an issue</a> · <a target="_blank" rel="ugc" href="https://discord.gg/RwVxY9Fuaj">Discord</a> · <a target="_blank" rel="ugc" href="https://www.youtube.com/@ausbossdev">YouTube @ausbossdev</a> · <a target="_blank" rel="ugc" href="https://x.com/Zanzibased">X @Zanzibased</a> · <a target="_blank" rel="ugc" href="https://github.com/ausboss">GitHub</a>

P-H3-Image Generate + Edit | 8 Pictures + Auto Prompt

Added a Size choice: 4 MP or 8 MP, for edits and new pictures. 4 MP is the default. For an edit it sets the picture loader's resize value. For a new picture, Latent Size picks the shape and the workflow scales it to that size. 8 MP gives faces and fine detail more room, with more memory and time, and P-H3 makes its best new pictures at 8 MP. It does not lock identity.
Low VRAM now uses the smaller prompt writer, p-h3-image_prompt_upsampler_w4a8 (6.7 GB instead of 19.3 GB). High VRAM keeps the full one. The prompts it writes came out nearly the same in my tests.

Files with SHA256: 69657db5c233315135302dc1c54e27c71e1e97ccae432a4a645171952a8799af