---
format: "aidr-story-markdown/v1"
id: "a87fce0409197d94559fd1d5c35ac4c11e1d090eefb831652e094d405dc38515"
canonical_url: "https://aidr.today/a87fce04?lang=en"
title: "fal's H3 Max Generates AI Video Faster Than Real Time for First Perpetual Stream"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-08-29T23:37:37.000Z"
category: "Models"
topics: ["minimax","video","open-source","inference","real-time","streaming"]
source_urls: ["https://huggingnews.com/ai/fals-h3-max-generates-ai-video-faster-than-real-time-for-first-perpetual-180be083","https://x.com/MiniMax_AI/status/2093778769903534294","https://x.com/MiniMax_AI/status/2093746422466367907","https://x.com/levelsio/status/2093628563693944889","https://x.com/levelsio/status/2093754163343593802","https://x.com/fal/status/2093844097148559588","https://x.com/fal/status/2093859932428812469","https://x.com/fal/status/2093844098977296474"]
summary: "fal developed H3 Max, a post-trained variant of MiniMax's open-weight H3 model, that generates 15 seconds of video in 9 seconds. The model operates 50x faster than the original H3 while maintaining quality, enabling AI video to be produced faster than it can be viewed. MiniMax said the achievement demonstrates the potential of open-weight models to advance latency and quality for interactive worlds and real-time storytelling. Developer levelsio used the model to create \"Infinite Slop,\" a live stream where user chat inputs are generated as sequential video in real time."
---

# fal's H3 Max Generates AI Video Faster Than Real Time for First Perpetual Stream

> [Open the canonical story](<https://aidr.today/a87fce04?lang=en>)

**Published:** 2026-08-29T23:37:37.000Z
**Category:** Models
**Topics:** minimax, video, open\-source, inference, real\-time, streaming

## Summary

fal developed H3 Max, a post\-trained variant of MiniMax's open\-weight H3 model, that generates 15 seconds of video in 9 seconds\. The model operates 50x faster than the original H3 while maintaining quality, enabling AI video to be produced faster than it can be viewed\. MiniMax said the achievement demonstrates the potential of open\-weight models to advance latency and quality for interactive worlds and real\-time storytelling\. Developer levelsio used the model to create "Infinite Slop," a live stream where user chat inputs are generated as sequential video in real time\.

## Sources

- [Story source](<https://huggingnews.com/ai/fals-h3-max-generates-ai-video-faster-than-real-time-for-first-perpetual-180be083>)
- [Story source](<https://x.com/MiniMax_AI/status/2093778769903534294>)
- [Supporting source](<https://x.com/MiniMax_AI/status/2093746422466367907>)
- [Supporting source](<https://x.com/levelsio/status/2093628563693944889>)
- [Story source](<https://x.com/levelsio/status/2093754163343593802>)
- [Story source](<https://x.com/fal/status/2093844097148559588>)
- [Story source](<https://x.com/fal/status/2093859932428812469>)
- [Supporting source](<https://x.com/fal/status/2093844098977296474>)