Pruna AI

Introducing P-Video-2-Pro-Replace, based on @MiniMax_AI H3! 🎬

Replace the people. Keep the scene.

Swap one or two people in an existing video using reference images, while preserving the original motion, camera, framing, background, and audio

And for launch, it's 50% off for the first two weeks, until November 19 at 6:00 PM CET!

Here's what you need to know:

- 𝗣𝗿𝗶𝗰𝗶𝗻𝗴: Starting from $0.05/s at 480p
- 𝗦𝗽𝗲𝗲𝗱: ~1.5–2 minutes for a 10-second video
- 𝗥𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲𝘀: Replace up to two people using reference images
- 𝗗𝘂𝗿𝗮𝘁𝗶𝗼𝗻: Supports videos up to 120 seconds

Available via our inference partners @runware, @wavespeed_ai, Sociaro, @TellersAI, @WiroAI, @Segmind_ai, @replicate, and @eachlabs

📚 Model docs: docs.api.pruna.ai/guides/models/…
🧩 Try the model through the API or Playground: dashboard.pruna.ai/login

P-Video-2-Pro-Replace is top-tier in overall human ratings, at a third of the cost and 1.6× faster.

- 𝐐𝐮𝐚𝐥𝐢𝐭𝐲: within 3% of the top-rated model at 480p
- 𝐂𝐨𝐬𝐭: 3.2× cheaper, just $0.05/s vs $0.16/s at 480p
- 𝐒𝐩𝐞𝐞𝐝: 1.6× faster than the next-fastest model, 6× faster than SkyReels V4
- #1 for 𝐔𝐆𝐂 ads at 768p, #1 and #2 for 𝐥𝐢𝐩 𝐬𝐲𝐧𝐜

Across both Quality × Price and Quality × Speed, P-Video-2-Pro-Replace stands out in the high-quality, low-cost, fast-generation region.

From UGC ads to music videos, swap characters without compromising on speed or budget.

Benchmarked by @Rapidata.

🧩 API & Playground: buff.ly/OHtx3eM
📚 Documentation: buff.ly/bHQHKV0
📊 P-Bench: buff.ly/Fk54Z2K

In October, our P-Creators used Pruna to bring their ideas to life and shared their creativity with all of us. Here are some of the highlights from this past month's Pruna Creative Program!

Thank you, P-Creators 💜

🎨 Start creating you too in the API or Playground: dashboard.pruna.ai/login
📚 Model docs: docs.api.pruna.ai/guides/models/…

Today, we open-source P-H3-Image, an image generation and editing model built from MiniMax-H3 by @MiniMax_AI. A SOTA video model, turned into an efficient open image model.

- 𝗛𝗼𝘄 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝘁 𝗶𝘀 𝗶𝘁? Generate or edit images 5× faster, in 8 steps instead of 50. A 4 MP image takes 2.6 s instead of 12.4 s. The model is 7× smaller than H3, so high-resolution images are a light download and run on a much smaller GPU.

- 𝗛𝗼𝘄 𝗴𝗼𝗼𝗱 𝗶𝘀 𝗶𝘁? Strongest at photorealism, documentary and editorial photography, film stills, and graphic styles with short quoted text. It renders 2 to 16 MP and edits with up to 8 input images.

-𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝗶𝘁 𝘄𝗼𝗿𝗸? We reworked every component of H3: the prompt upsampler, the text encoder, the diffusion schedule, the transformer, and the decoder. Use it with Diffusers or ComfyUI.

📄 Tech Report: pruna.ai/blog/oss-p-h3-…
🤗 @huggingface Open Weights: huggingface.co/PrunaAI/P-H3-I…
👉 @ComfyUI Native Support: huggingface.co/PrunaAI/P-H3-I…

The P-Video-2-Pro family just got a resolution upgrade! 🎬

We've added 1080p, 2K, and 4K output to all three P-Video-2-Pro models:

• P-Video-2-Pro
• P-Video-2-Pro-Reference
• P-Video-2-Pro-Avatar

Generate videos in higher resolutions, directly through our API or Playground!

📚 Model docs: docs.api.pruna.ai
🧩 Try it out: dashboard.pruna.ai/login