Haiper, the AI platform for visual content creation, this week launched its new Haiper 2.0 model, to help users generate hyper-realistic videos and images faster than ever before, setting a new industry standard for quality.
Haiper 2.0 utilises a proprietary combination of transformer-based models and diffusion techniques, significantly enhancing the quality, realism, and creation speed of generated content. Haiper recently surpassed the 4.5 million user milestone in its first seven months out of stealth. Alongside this updated model, Haiper is also launching Video Templates—blueprint formats for creative, trending videos or product marketing ads. The categories of Templates continue to grow, as Haiper's community contributes new designs.
The new Haiper 2.0 model will produce higher quality outputs, 4K resolution soon, and faster generation times. The new model will also increase temporal coherence with smoother and more lifelike movements. As Haiper scales, the company’s perceptual diffusion transformer (DiT) video models and advanced neural architectures will provide unparalleled performance in resolution, dynamism, and speed.
Dr Yishu Miao, co-founder and CEO of Haiper, commented: “Haiper 2.0 is a groundbreaking upgrade to our widely used and celebrated video generation model. Users will be able to generate ultra-realistic videos faster than ever before. At Haiper, we incorporate creators’ feedback to actively improve aspects that matter most to them: generation speed, realism, and temporal consistency. Launching Templates alongside Haiper 2.0 will make it easier for creators—whether hobbyists, enthusiasts or enterprise customers—to explore new use cases.”
Launched: Video Templates
Haiper 2.0 launches alongside Templates, a series of well-designed workflows to help users animate and transform images into videos. Templates include a range of business use cases, such as logo and product animations; and fun, viral videos: users can morph still images of people and backgrounds into shareable videos, such as characters kissing or hugging, face swapping, or viral TikTok dances. Users simply upload images to a template, and Haiper generates a high-quality video, saving time on typing specific prompts. Videos can then be adapted with further prompts, to ensure generations precisely fit the user’s vision.
Haiper History
Haiper was founded in late 2021 by Dr. Yishu Miao (CEO) and Dr. Ziyu Wang (CTO), both with PhDs in machine learning from Oxford University and former researchers at Google’s DeepMind. The company’s co-founders have experience working alongside AI pioneers, including Geoffrey Hinton, Nando de Freitas, and Phil Blunsom.
Since Haiper’s launch and $13.8M seed round earlier in 2024, the platform has launched additional capabilities, including HD upscaler, text-to-image generation, and keyframe conditioning for ultra-precise videos. Haiper is also leveraging NVIDIA technologies as they explore video generation up to 30 seconds.
About Haiper
Haiper is a video-generative AI platform building a powerful foundation model for visual content. Scaling to become Europe’s powerhouse for visual generative AI, the company has raised $19.2M to date. Founded in late 2021 by Dr. Yishu Miao (CEO) and Dr. Ziyu Wang (CTO), Haiper has expanded its team to 30 people, from eight different countries. The company is headquartered in London and also has an office in Toronto. To learn more, visit https://haiper.ai.
About Haiper
Haiper is a video-generative AI platform building a powerful foundation model for visual content. Scaling to become Europe’s powerhouse for visual generative AI, the company has raised $19.2M to date. Founded in late 2021 by Dr. Yishu Miao (CEO) and Dr. Ziyu Wang (CTO), Haiper has expanded its team to 30 people, from eight different countries. The company is headquartered in London and also has an office in Toronto. To learn more, visit https://haiper.ai.
