diff --git a/README.md b/README.md
index ca2f1fe1edb2c1be022ea23375ace151f0e73fae..ca2f4c4d43fce81a42d0cd4aadd18b01f139c298 100644
--- a/README.md
+++ b/README.md
@@ -28,27 +28,16 @@ tags:
- 
- 
- 
+
+
-
# MiniMax H3
-## Online API
-Use MiniMax\-H3 directly via API\.
-- Global: [platform\.minimax\.io](https://platform.minimax.io/docs/api-reference/video-generation-v2-create) \| CN: [platform\.minimaxi\.com](https://platform.minimaxi.com/docs/api-reference/video-generation-v2-create)
-
-## Online App
-Use MiniMax\-H3 directly via App\.
-- WebApp Global: [hailuoai\.video](https://hailuoai.video/tools/minimax-h3) \| CN: [hailuoai\.com](https://hailuoai.com/)
-- Desktop Global: [hub\.minimax\.io](https://hub.minimax.io/) \| CN: [hub\.minimaxi\.com](https://hub.minimaxi.com/)
-
## System Overview
MiniMax H3 is a general-purpose, omni-modal generative system. It supports unified understanding of multimodal contexts composed of text, images, video, and audio, and can generate video with native stereo audio at resolutions up to 2K and durations of up to 15 seconds. Thanks to its task-generalization-oriented system design, H3 already possesses broad multimodal context understanding and generation capabilities at the pre-training stage, enabling outstanding performance in following complex multimodal instructions.
@@ -77,6 +66,20 @@ The complete H3 system consists of the following three modules:
- H3-Base: Generates audio and video based on the H3-Context-IR output, producing results at 768p resolution.
- H3-Regenerate-2K: Feeds the 768p result together with the original context back into H3 to regenerate the output at 2K resolution. This process leverages both H3’s powerful generative capabilities and the rich information contained in the original context, enabling it to produce high-resolution outputs with more accurate details and greater visual fidelity.
+## Online API
+
+Use MiniMax\-H3 directly via API\.
+
+- Global: [platform\.minimax\.io](https://platform.minimax.io/docs/api-reference/video-generation-v2-create) \| CN: [platform\.minimaxi\.com](https://platform.minimaxi.com/docs/api-reference/video-generation-v2-create)
+
+## Online App
+
+Use MiniMax\-H3 directly via App\.
+
+- WebApp Global: [hailuoai\.video](https://hailuoai.video) \| CN: [hailuoai\.com](https://hailuoai.com/)
+
+- Desktop Global: [hub\.minimax\.io](https://hub.minimax.io/) \| CN: [hub\.minimaxi\.com](https://hub.minimaxi.com/)
+
## Model Architecture
### H3\-Context\-IR
@@ -189,27 +192,19 @@ Each checkpoint is distributed as a self\-contained Hugging Face\-style reposito
└── audio_vae/
```
-Download the model. The repository hosts the original checkpoint (`FL2VA/`, `Ref2VA/`) and the diffusers format side by side, so scope the download to what your framework needs:
-
-`model_index.json` is the repository-level public entry. The task-family-specific diffusers indexes remain under `FL2VA/model_index.json` and `Ref2VA/model_index.json`.
+Download the model:
```bash
-# Original checkpoint, both task families (SGLang, vLLM):
-hf download MiniMaxAI/MiniMax-H3 --include "model_index.json" "FL2VA/*" "Ref2VA/*" --local-dir MiniMax-H3
-
-# Or a single task family:
-hf download MiniMaxAI/MiniMax-H3 --include "model_index.json" "FL2VA/*" --local-dir MiniMax-H3
+hf download MiniMaxAI/MiniMax-H3 --local-dir MiniMax-H3
```
-diffusers users do not need a manual download: `ModularPipeline.from_pretrained("MiniMaxAI/MiniMax-H3")` fetches exactly the components it needs. See the [diffusers documentation](https://huggingface.co/docs/diffusers/main/en/api/pipelines/minimax_h3) for loading recipes.
-
We recommend the following inference frameworks to serve the model:
- [SGLang](https://docs.sglang.io/) \- see [cookbook](https://docs.sglang.io/cookbook/diffusion/MiniMax/MiniMax-H3)
- [vLLM](https://github.com/vllm-project/vllm) \- see [vllm recipes](https://recipes.vllm.ai/MiniMaxAI/MiniMax-H3)
-- [diffusers](https://github.com/huggingface/diffusers) \- see [diffusers docs](https://huggingface.co/docs/diffusers/main/en/api/pipelines/minimax_h3)
+- [diffusers](https://github.com/huggingface/diffusers) \- see [diffusers docs](https://huggingface.co/docs/diffusers/en/api/models/autoencoderkl_minimax_h3)
- [ComfyUI](https://github.com/Comfy-Org/ComfyUI) \- see [Comfy tutorial](https://docs.comfy.org/tutorials/video/minimax/minimax-h3); use [R2V template](https://github.com/Comfy-Org/workflow_templates/blob/main/templates/video_minimax_h3_r2v.json) / [T2V template](https://github.com/Comfy-Org/workflow_templates/blob/main/templates/video_minimax_h3_t2v.json)
@@ -274,11 +269,8 @@ TOKEN=""
MiniMax platform:
-API docs:
-- Create H3-2K: use /video-generation-v2-create [EN-docs](https://platform.minimax.io/docs/api-reference/video-generation-v2-create), [CN-docs](https://platform.minimaxi.com/docs/api-reference/video-generation-v2-create)
-- H3-Context-IR：use /video-generation-v2-h3-context-ir [EN-docs](https://platform.minimax.io/docs/api-reference/video-generation-v2-h3-context-ir), [CN-docs](https://platform.minimaxi.com/docs/api-reference/video-generation-v2-h3-context-ir)
-- H3-Regenerate-2K：use /video-generation-v2-regeneration [EN-docs](https://platform.minimax.io/docs/api-reference/video-generation-v2-regeneration), [CN-docs](https://platform.minimaxi.com/docs/api-reference/video-generation-v2-regeneration)
-
+- CN: https://platform.minimaxi.com
+- Global: https://platform.minimax.io
The examples below encode local H3\-Base output files as Base64 Data URLs\. For production use, uploading the video to a publicly accessible URL and passing that URL as `base_video` is recommended\.
@@ -407,13 +399,11 @@ For each case below, we provide reference outputs at both 2K and 768p generated
[VIDEO\_PROMPT\_WRITING\_GUIDE\_ref\_en\.md](docs/VIDEO_PROMPT_WRITING_GUIDE_ref_en.md)
-skills to improve prompt: https://github.com/MiniMax-AI/MiniMax-H3/tree/main/skills
+
## License
-- MiniMax H3 is released under the [MiniMax H3 Community License Agreement](https://huggingface.co/MiniMaxAI/MiniMax-H3/blob/main/LICENSE).
-- [Q&A about the License](docs/QA-about-License.md)
-- [Application form(only for USA/EU/UK/South Korea)](https://platform.minimax.io/h3-license)
+MiniMax H3 is released under the [MiniMax H3 Community License Agreement](LICENSE). [Q&A about License](docs/QA-about-License.md)
## Contact Us
diff --git a/assets/action-reference.mov b/assets/action-reference.mov
new file mode 100644
index 0000000000000000000000000000000000000000..1b69b6924dccb629703fed760b10ab0e20c4b7a1
--- /dev/null
+++ b/assets/action-reference.mov
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:116e0f08a399834e7ffc3472d036659b33250f4ba4f0b7e63f17ec07cb58e4dc
+size 31446151
diff --git a/assets/character-action-reference.png b/assets/character-action-reference.png
new file mode 100644
index 0000000000000000000000000000000000000000..b4023d56c08bf70b44383a82bacbacd837a682e4
--- /dev/null
+++ b/assets/character-action-reference.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01a6c6f29e1ce9d414276e6e2eca06b171e5a68dd54a86074a4ab77bb9cb8883
+size 5239867
diff --git a/assets/character-replacement-action-reference.mp4 b/assets/character-replacement-action-reference.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3fbb5cf0d174c34bf858a05f2b7b4ba74c8d2083
--- /dev/null
+++ b/assets/character-replacement-action-reference.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:796c264126110d87992dcb54213ac0697920cb4ddf3d5a06aa36069386fc4fa1
+size 8223683
diff --git a/assets/fashion-glasses-ad.mp4 b/assets/fashion-glasses-ad.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b4981e45a1e1fb3187480a7dce633ad960aa1615
--- /dev/null
+++ b/assets/fashion-glasses-ad.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a11e6b0095dd7766b724bb886edf4d7d9930af44da6839ee195269ac4fc60ba4
+size 10938085
diff --git a/assets/fashion-glasses-reference-1.png b/assets/fashion-glasses-reference-1.png
new file mode 100644
index 0000000000000000000000000000000000000000..b263a76e90c26b3bb8a5d57ea1f20f0deca687df
--- /dev/null
+++ b/assets/fashion-glasses-reference-1.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5665f6529dd73339c0ac9397f0bd52df0cd44bd5b63d68af88f95855879d3d8b
+size 1008875
diff --git a/assets/fashion-glasses-reference-2.png b/assets/fashion-glasses-reference-2.png
new file mode 100644
index 0000000000000000000000000000000000000000..1b70cbb225fbd4318f338f25c71e9961a074b216
--- /dev/null
+++ b/assets/fashion-glasses-reference-2.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a74bdab0bff0e126aae3757833fd2ec8cfba9afcb2ca36f63ee86977f7f7c00
+size 889303
diff --git a/assets/fashion-glasses-reference-3.png b/assets/fashion-glasses-reference-3.png
new file mode 100644
index 0000000000000000000000000000000000000000..12f4213eba0fe1da2768fdfc4c6f5f51df44a9ce
--- /dev/null
+++ b/assets/fashion-glasses-reference-3.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b1491e5a227f226fa1395e71ac89a84358a435d974acb9f9065b9ded0982cbe
+size 1057045
diff --git a/assets/fashion-glasses-reference-4.png b/assets/fashion-glasses-reference-4.png
new file mode 100644
index 0000000000000000000000000000000000000000..3fa4966da9917d0763efcc23ba8ffa0a03c3bad6
--- /dev/null
+++ b/assets/fashion-glasses-reference-4.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9db6d6788855029ef9ac4109b05e17693ddb3bdf6ae3f24cec44de478c00f13
+size 368310
diff --git a/assets/fl2va-clay-fox-reference.png b/assets/fl2va-clay-fox-reference.png
new file mode 100644
index 0000000000000000000000000000000000000000..a2a9c1c51aeaeb566598640e3489d903eb4b2c3c
--- /dev/null
+++ b/assets/fl2va-clay-fox-reference.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55eff74469b79a100e63bdd358596f6cff671e18fca9b276d6faef16faf22d27
+size 1732646
diff --git a/assets/fl2va-clay-fox.mp4 b/assets/fl2va-clay-fox.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..286afec6672ac6b396f2470643661b956dd70f30
--- /dev/null
+++ b/assets/fl2va-clay-fox.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4604b0838b736ecaa209c978f880bfacbb3a0fc82a8a0517c1f5aa16454b50ef
+size 10931936
diff --git a/assets/h3-architecture.png b/assets/h3-architecture.png
new file mode 100644
index 0000000000000000000000000000000000000000..1d09e39a296a719a9d470286b08257016774fe0e
--- /dev/null
+++ b/assets/h3-architecture.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9765612d331f5fd31068a9283ffe28f11963f6a418c9f4093f81b408ea756630
+size 335879
diff --git a/assets/h3-cinematic-shot.mp4 b/assets/h3-cinematic-shot.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4ca80a1ce5db389f0d2d0c1376c3830824893e0a
--- /dev/null
+++ b/assets/h3-cinematic-shot.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:436defc81cfa7d53aef423f368be82ead20056088087e0be308b6e3865e8fb81
+size 4361547
diff --git a/assets/h3-suspense-title.mp4 b/assets/h3-suspense-title.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a6f6a14bb3d2a203f7750b42fe3912304ac2b361
--- /dev/null
+++ b/assets/h3-suspense-title.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90ebbd7edc71c9a0151c3064126acd2dbb5b7cba58ce459e223e29f2c04f9186
+size 16809220
diff --git a/assets/logo.svg b/assets/logo.svg
new file mode 100644
index 0000000000000000000000000000000000000000..972b41cdb04e1ab4c2b94fa0c35df0c80a5151a2
--- /dev/null
+++ b/assets/logo.svg
@@ -0,0 +1,16 @@
+
diff --git a/assets/reference-image-1.png b/assets/reference-image-1.png
new file mode 100644
index 0000000000000000000000000000000000000000..5162fc9cd50c4f5dfdde01e0a55c65fad25e2962
--- /dev/null
+++ b/assets/reference-image-1.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49b89115da4e253228969830ca491e4ca1b79331eb241a905d1ebca3dc6de7be
+size 720511
diff --git a/assets/reference-image-2.png b/assets/reference-image-2.png
new file mode 100644
index 0000000000000000000000000000000000000000..3fa4966da9917d0763efcc23ba8ffa0a03c3bad6
--- /dev/null
+++ b/assets/reference-image-2.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9db6d6788855029ef9ac4109b05e17693ddb3bdf6ae3f24cec44de478c00f13
+size 368310
diff --git a/assets/robot-arm-red-cube.mp4 b/assets/robot-arm-red-cube.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f731941695bf691022996a50020d3d86c0ffed3f
--- /dev/null
+++ b/assets/robot-arm-red-cube.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a751e6100dbf6502a99a2adc0b12171304ca7b17eecea70e2ed035a73ea692c
+size 2259732
diff --git a/assets/t2va-768p-demo.mp4 b/assets/t2va-768p-demo.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..51d3b9b7c285efa8b1c7c168a6e5ced73468bbad
--- /dev/null
+++ b/assets/t2va-768p-demo.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d66903241362e224085cc93f7a5e70fba6ab378d0ac6ff6af83acf2559849a42
+size 1637373
diff --git a/docs/QA-about-License.md b/docs/QA-about-License.md
index f31bbf62d5011644a9470537d171ed8adb10863b..08ad0d3b9e61d74a278fe1ac81d6251e198aee7d 100644
--- a/docs/QA-about-License.md
+++ b/docs/QA-about-License.md
@@ -45,7 +45,7 @@ Yes.
Organizations in these regions can apply for a formal license. After reviewing the deployment scenario and confirming that appropriate compliance controls and safeguards are implemented, MiniMax may authorize usage.
-[Application form](https://platform.minimax.io/h3-license)
+[Application form](https://vrfi1sk8a0.feishu.cn/share/base/form/shrcnD9XM1zYI9VFJxTEbt0d19g)
Through authorized deployments, MiniMax can ensure that MiniMax-H3 is used responsibly while meeting local legal and regulatory requirements.
diff --git a/model_index.json b/model_index.json
deleted file mode 100644
index 454be19164878173a962c39c81633e4377880fca..0000000000000000000000000000000000000000
--- a/model_index.json
+++ /dev/null
@@ -1,131 +0,0 @@
-{
- "_class_name": "MiniMaxH3ModularPipeline",
- "_diffusers_version": "0.36.0.dev0",
- "_blocks_class_name": "MiniMaxH3Blocks",
- "text_encoder": [
- "transformers",
- "Qwen3VLForConditionalGeneration",
- {
- "type_hint": [
- "transformers",
- "Qwen3VLForConditionalGeneration"
- ],
- "pretrained_model_name_or_path": "MiniMaxAI/MiniMax-H3",
- "subfolder": "text_encoder",
- "variant": null,
- "revision": null
- }
- ],
- "tokenizer": [
- "transformers",
- "Qwen2TokenizerFast",
- {
- "type_hint": [
- "transformers",
- "Qwen2TokenizerFast"
- ],
- "pretrained_model_name_or_path": "MiniMaxAI/MiniMax-H3",
- "subfolder": "tokenizer",
- "variant": null,
- "revision": null
- }
- ],
- "processor": [
- "transformers",
- "Qwen3VLProcessor",
- {
- "type_hint": [
- "transformers",
- "Qwen3VLProcessor"
- ],
- "pretrained_model_name_or_path": "MiniMaxAI/MiniMax-H3",
- "subfolder": "processor",
- "variant": null,
- "revision": null
- }
- ],
- "vae": [
- "diffusers",
- "AutoencoderKLMiniMaxH3",
- {
- "type_hint": [
- "diffusers",
- "AutoencoderKLMiniMaxH3"
- ],
- "pretrained_model_name_or_path": "MiniMaxAI/MiniMax-H3",
- "subfolder": "vae",
- "variant": null,
- "revision": null
- }
- ],
- "audio_vae": [
- "diffusers",
- "AutoencoderKLMiniMaxH3Audio",
- {
- "type_hint": [
- "diffusers",
- "AutoencoderKLMiniMaxH3Audio"
- ],
- "pretrained_model_name_or_path": "MiniMaxAI/MiniMax-H3",
- "subfolder": "audio_vae",
- "variant": null,
- "revision": null
- }
- ],
- "transformer": [
- "diffusers",
- "MiniMaxH3Transformer3DModel",
- {
- "type_hint": [
- "diffusers",
- "MiniMaxH3Transformer3DModel"
- ],
- "pretrained_model_name_or_path": "MiniMaxAI/MiniMax-H3",
- "subfolder": "transformer",
- "variant": null,
- "revision": null
- }
- ],
- "transformer_ref": [
- "diffusers",
- "MiniMaxH3Transformer3DModel",
- {
- "type_hint": [
- "diffusers",
- "MiniMaxH3Transformer3DModel"
- ],
- "pretrained_model_name_or_path": "MiniMaxAI/MiniMax-H3",
- "subfolder": "transformer_ref",
- "variant": null,
- "revision": null
- }
- ],
- "scheduler": [
- "diffusers",
- "MiniMaxH3Scheduler",
- {
- "type_hint": [
- "diffusers",
- "MiniMaxH3Scheduler"
- ],
- "pretrained_model_name_or_path": "MiniMaxAI/MiniMax-H3",
- "subfolder": "scheduler",
- "variant": null,
- "revision": null
- }
- ],
- "audio_scheduler": [
- "diffusers",
- "MiniMaxH3Scheduler",
- {
- "type_hint": [
- "diffusers",
- "MiniMaxH3Scheduler"
- ],
- "pretrained_model_name_or_path": "MiniMaxAI/MiniMax-H3",
- "subfolder": "audio_scheduler",
- "variant": null,
- "revision": null
- }
- ]
-}