Hailuo 2.3
Neural Network
Hailuo 2.3 is a MiniMax video model for generating video from text descriptions or reference images.
Max answer length
(in tokens)
Context size
(in tokens)
Video generation
(duration 8 sec., 720p)
Video generation
(duration 8 sec., 1080p)
How it works Hailuo 2.3?
Hailuo 2.3 is a generative video model by MiniMax. It accepts text prompts and reference images, operating in two modes: text-to-video, where clips are created from text descriptions, and image-to-video, where motion and scenes are built from a provided frame. The reference helps maintain the desired composition, style, and character appearance without needing verbal descriptions. Via BotHub, the model is accessible from Russia without a VPN or foreign card: pay with a Russian card in rubles only for what you generate, with no subscriptions or expiring tokens. Over 250 models are available in the same interface, allowing you to combine video, text, and voiceover in one window. For business tasks, there is a unified OpenAI-compatible API and corporate access via contract with invoicing and an admin panel. Use cases include: advertising and product videos from product photos, short vertical videos for social media, animation of illustrations and concept art, storyboards and animatics for clips, and quick visual prototypes of ideas before filming.Frequently asked questions about Hailuo 2.3
You can use generated results for commercial purposes. You own all rights to the content you create. The only restriction: make sure your prompt does not include copyrighted third-party material. You are responsible for respecting the rights to any input data.
hailuo-2.3 by MiniMax creates videos: from text descriptions or an uploaded image that becomes a reference frame. Suitable for short clips, advertising concepts, animation of illustrations, scenes with camera movement, and character dynamics. Accessible from Russia without a VPN or foreign card, payment in rubles.
You receive a ready-to-use video file that can be downloaded and immediately edited. Exact duration and resolution parameters are specified in the technical header of the page. The quality of the result depends heavily on the detail of the scene description and the quality of the source image when generating from a frame.
This is a generative video model, and there is no specific note about step-by-step reasoning before answering in our data. It is better to plan the scene on your side: describe the frame, camera movement, light, and action. If you need reasoning chains, switch to a text model in BotHub.
In our data, function calling and structured JSON output are not noted for hailuo-2.3: the model's output is a video file, not a text structure. However, you can access it via the unified BotHub OpenAI-compatible API using the same key as for other models.
Images — yes: upload a frame, and the model will animate it based on the composition and style of the reference. There are no notes about accepting audio or video input in our data, so we will not promise such a scenario. A text prompt works on its own, without an image.
We have no data on how the model works with Russian prompts, so we won't promise anything — test it with your task. For video, visual details matter: object, action, shot, light, camera movement. The BotHub interface and support are Russian, with payment in rubles without a foreign card.