Hailuo / MiniMax Video-01
MiniMax's Hailuo video-01. 6s 1280x720 clips with strong cinematic motion and physical realism.
Hailuo / MiniMax Video-01 is video generation AI model from Custom, priced at €0.000 per 1M input tokens with a unknown context window.
Pricing
API Integration
Use our OpenAI-compatible API to integrate Hailuo / MiniMax Video-01 into your application.
npm install railwailimport railwail from "railwail";
const rw = railwail("YOUR_API_KEY");
// Simple — just pass a string
const reply = await rw.run("hailuo-video-01", "Hello! What can you do?");
console.log(reply);
// With message history
const reply2 = await rw.run("hailuo-video-01", [
{ role: "system", content: "You are a helpful assistant." },
{ role: "user", content: "Explain quantum computing simply." },
]);
console.log(reply2);
// Full response with usage info
const res = await rw.chat("hailuo-video-01", [
{ role: "user", content: "Hello!" },
], { temperature: 0.7, max_tokens: 500 });
console.log(res.choices[0].message.content);
console.log(res.usage);Deep dive — MiniMax's Hailuo / MiniMax Video-01
MiniMax (Shanghai Xiyu Technology) was founded in 2021 by Yan Junjie and a team of former SenseTime and Microsoft Research Asia researchers. It is one of China's 'four AI tigers' and has raised over $1B from Tencent, Alibaba and HongShan, reaching a $2.5B+ valuation in 2024. MiniMax built the abab LLM family, the Hailuo AI consumer assistant and the Hailuo Video product, the latter launched in September 2024 with Hailuo Video-01 as the first publicly available text-to-video and image-to-video model. The model immediately drew attention for its cinematic motion and strong character consistency on image-to-video, becoming the top-ranked Chinese model on the public Artificial Analysis video leaderboard at launch.
Visit MiniMax →Hailuo Video-01 is the first generation of MiniMax's closed-source video diffusion model. It is a transformer-based diffusion model operating on a learned spatio-temporal latent space, conditioned on bilingual Chinese/English text embeddings and optional image embeddings (first frame or character reference). MiniMax publish no architectural details, but the model generates 6-second clips at 1280x720 / 25 fps with optional 1080p output. It supports text-to-video, image-to-video and subject-reference modes. Training presumably uses a large curated multilingual video corpus with synthetic dense captions; post-training likely combines a reward-model objective for aesthetics with caption-alignment supervision. Subsequent versions (Video-01-Live, Video-01-Subject) added live2D-style animation of stylised characters and subject locking.
- Parameters
- Undisclosed
- Context
- unknown
- Text-to-video and image-to-video generation at 720p-1080p
- 6-second clips at 25 fps
- Strong character consistency via subject-reference mode
- Bilingual Chinese/English prompting
- Live2D-style anime animation in Video-01-Live variant
- Available via Hailuo AI web/mobile and API
- Director-style camera control vocabulary
- Competitive pricing for high resolution
- Best for: anime/stylised animation, character shorts, social-media reels.
Closed corpus of licensed and web video with bilingual dense captions; exact size undisclosed.
License: Proprietary commercial licence via Hailuo AI / MiniMax API.
Known limitations
- 6-second clip limit
- No native audio
- Heavy political and public-figure moderation
- Less prompt-following precision than Veo 3 / Sora 2
- Closed model, no technical paper
Research papers
Frequently asked questions
Related Models
View all Video GenerationGoogle Veo 2
Google's state-of-the-art video generation model. Simulates real-world physics with various visual styles.
Google Veo 3
Google's Veo 3. High-fidelity text-to-video with native audio generation, up to 8s clips.
Google Veo 3.1
Latest Veo with image-to-video and context-aware audio
Kling v3
Cinematic video up to 15s with multi-shot and native audio
Start using Hailuo / MiniMax Video-01 today
Get started with free credits. No credit card required. Access Hailuo / MiniMax Video-01 and 100+ other models through a single API.