<p><a target="_blank" rel="ugc" href="https://hailuoai.video/"><strong>Hailuo H3</strong></a> is the newest video generation model from <a target="_blank" rel="ugc" href="https://www.minimax.io/">MiniMax</a>, the Shanghai AI lab behind the Hailuo video app and the MiniMax text and audio model families. H3 generates native 2K video with synchronized audio in a single pass, at clip lengths up to 15 seconds.</p><p><br /><strong>Developed by MiniMax. </strong>All credit for the model goes to the MiniMax team. You can use H3 two ways on Civitai: generate on-site against the hosted <a target="_blank" rel="ugc" href="https://platform.minimax.io/docs/guides/video-generation">MiniMax API</a> (the H3 (Api) version), or download the open weights from the H3 (Comfy) version on this page and run it yourself. The ComfyUI-format weights are also available from MiniMax and Comfy-Org on Hugging Face.</p><p></p><h3 id="background">Background</h3><p>MiniMax has shipped video models under the Hailuo name since 2024, moving from the I2V-01 series through Hailuo 02 and Hailuo 2.3. Those earlier models topped out around 1080p with fixed 6 or 10 second clips and no native sound. H3 is the step past that: it raises the ceiling to 2K, opens the clip length to any integer from 4 to 15 seconds, and folds audio generation into the same pass as the picture rather than bolting it on afterward.</p><p></p><h3 id="capabilities">Capabilities</h3><p>• Native 2K output (2560x1440 at 16:9)<br />• Clips from 4 to 15 seconds, in one second increments<br />• Synchronized audio generated with the video - dialogue, effects, and ambience timed to what is on screen<br />• Text-to-video, image-to-video, and first frame plus last frame control<br />• Omni-reference conditioning: up to 9 reference images, 3 reference video clips, and 3 reference audio clips<br />• Multi-shot sequences and instruction-based editing</p><p></p><h3 id="on-civitai">On Civitai</h3><p>H3 is wired into the video generator across four workflows: <strong>Create Video</strong> (text only), <strong>Image to Video</strong> (single first frame), <strong>First/Last Frame</strong> (both endpoints), and <strong>Reference to Video</strong> (up to 9 reference images). Resolution is fixed at 2K because that is the only output H3 produces. Aspect ratio is selectable at 21:9, 16:9, 4:3, 1:1, 3:4, and 9:16 for text prompts; when you supply a frame image, H3 adapts the framing to that image instead.</p><p>Reference video and reference audio inputs are supported by the model but are not exposed in the Civitai form yet.</p><p></p><h3 id="notes">Notes</h3><p>H3 is a new model and MiniMax has not published a full spec sheet or a stable pricing page for it. Behavior and cost may shift as they finish the rollout. Prompt handling follows the usual MiniMax posture - this is a commercially hosted service and content is subject to the provider's policies, so expect refusals on material their API declines.</p><p></p><h3 id="links">Links</h3><p>• <a target="_blank" rel="ugc" href="https://platform.minimax.io/docs/guides/video-generation">MiniMax video generation docs</a><br />• <a target="_blank" rel="ugc" href="https://hailuoai.video/">Hailuo AI</a><br />• <a target="_blank" rel="ugc" href="https://www.minimax.io/">MiniMax</a><br />• <a target="_blank" rel="ugc" href="https://hailuoai.video/doc/terms-of-service.html">MiniMax Hailuo AI Terms of Service</a></p>

MiniMax H3

minimaxH3_h3Comfy_3085953.safetensors

minimax_h3_video_vae_fp16.safetensors

minimaxH3_h3Comfy_3085597.safetensors

minimaxH3_h3Comfy_nvfp4.safetensors

minimaxH3_h3Comfy_3085684.safetensors

minimaxH3_h3Comfy_3086249.safetensors

minimaxH3_h3Comfy_3086349.safetensors

minimaxH3_h3Comfy_3086408.safetensors

minimax_h3_audio_vae_fp32.safetensors

minimaxH3_h3Comfy_3098340.safetensors

minimaxH3_h3Comfy_3098344.safetensors

minimaxH3_h3Comfy_3098338.safetensors

minimaxH3_h3Comfy_3085569.safetensors

minimaxH3_h3Comfy_3085784.safetensors

H3 (Comfy)

H3 (Api)

Files with SHA256: bc2ced0fbea64757fa9acddccfc0b3f4819d1dcf1da6c124d690d368be283923