MiniMax H3 and H3 Turbo video

Sogni runs the open-weight MiniMax H3 FL2VA model through ComfyUI for text-to-video, first-frame image-to-video, and first/last-frame video, plus the separate MiniMax H3 Ref2VA checkpoint for multi-reference video. MiniMax H3 Turbo applies LightX2V's 4-step distillation LoRA to the three FL2VA workflows for about 3.5× faster iteration. Video and synchronized stereo audio are generated together in one pass. Sogni has written authorization from MiniMax to offer MiniMax H3 through the Sogni platform.

MiniMax H3 is a separate MiniMax model, not a Seedance derivative. Its short-form multimodal scope is comparable to Seedance 2.0, while the Sogni deployment uses released H3 weights on Sogni Supernet workers.

#Workflow IDs

Turbo currently covers FL2VA text-to-video, image-to-video, and first/last-frame video. There is no Turbo Ref2VA model ID.

#API and Creative Agent availability

All seven current modes in the table are available through both the Sogni API and the Sogni Creative Agent Skill: the four standard H3 workflows (text-to-video, image-to-video, first/last-frame video, and Ref2VA) and the three Turbo FL2VA workflows. The API accepts the exact model IDs above. Creative Agent selects the matching workflow from your request and supplied media, including labelled image, video, and audio references for standard Ref2VA.

Sogni Web currently exposes standard text-to-video, image-to-video, and Ref2VA. Use the API or Creative Agent Skill for standard first/last-frame generation and the three Turbo modes.

#Standard or Turbo?

Choose H3 Turbo for drafts, prompt iteration, timing tests, and other work where turnaround matters most. Turbo is also about 2.67× cheaper, billing 6 Spark per second against standard H3's 16 — see Pricing. Sogni's warm reference runs measured about 3.5× faster for image-to-video and 3.7× faster for text-to-video, conservatively summarized as about 3.5× faster; actual wall-clock time varies with mode, input conditioning, clip length, resolution, worker hardware, and queue state.

Choose standard 20-step H3 when fine visual detail and audio polish matter more than iteration speed. LightX2V labels the current v0.1 Turbo weights a preview and notes that image detail still needs improvement. Turbo is a speed/quality trade-off, not a higher-quality replacement for standard H3.

On Sogni's primary NVIDIA RTX PRO 6000 H3 workers, a highest-quality 15-second standard H3 video typically takes about 15 minutes once rendering starts; H3 Turbo typically takes about 5 minutes. Raw render time is the same on Unlimited and Unlimited Pro. Pro changes how much work can run and where it sits in the subscription queue, not the speed of one GPU render. Sogni is testing current community workflow optimizations that may improve these times by up to 20%.

#Generation limits

• Fixed 24 fps with native 32 kHz stereo audio.
• 124–362 frames on the 124 + n×17 frame grid, approximately 5.17–15.08 seconds.
• Current output is 768p-class. Dimensions use a 32 px grid, up to 1344 px per side and no more than 1,032,192 rendered pixels. The primary landscape and portrait sizes are 1344×768 and 768×1344.
• Standard H3 uses the fixed 20-step res_multistep / simple workflow. H3 Turbo uses its fixed 4-step distilled sampling path. Guidance is fixed at 1 for both.
• Ref2VA reference sets: up to 9 images, 3 videos (24 fps, 2–15 s, each with an optional soundtrack), and 3 standalone audio clips — at most 12 reference files, with at least one visual reference (image or video). Audio alone is rejected.

#Pricing

MiniMax H3 and H3 Turbo are priced per second of finished video, at a flat rate for each speed tier at the standard 1 Spark = $0.005 peg. Standard H3 bills 16 Spark per second, which is $0.08 per second of render value — the same 768p rate as the official MiniMax API. H3 Turbo bills 6 Spark per second, which is $0.03 per second — about 2.67× cheaper than standard. Within a tier, duration is the only input that changes the price.

• Resolution and aspect ratio are not price factors. A square 768×768 clip and a widescreen 1344×768 clip of the same length cost exactly the same.
• Mode is not a price factor, but speed tier is. Standard text to video, image to video, first/last frame, and Ref2VA reference to video all share the 16 Spark per second standard rate; Turbo text to video, image to video, and first/last frame all share the 6 Spark per second Turbo rate.
• References are free. Reference images, videos, and audio carry no surcharge, at any count.
• You pay for the length that actually renders. H3 snaps to the 124 + n×17 frame grid at 24 fps, so billable durations land on grid values between 5.17 s and 15.08 s.

H3 runs on Sogni Supernet workers rather than an external vendor API, so it uses standard Spark access and does not require credit-card Premium Spark coverage. For a live quote at any frame count, use the job estimate calculator.

#Unlimited throughput and fair use

Both Sogni Unlimited plans cover standard H3 and H3 Turbo under credit-free fair use. The render time for one job is the same on both plans, but Unlimited Pro adds higher subscription queue priority, an additional standard H3 concurrency slot, and twice the H3 fair-use capacity of Unlimited.

Sogni intentionally does not publish a fixed number of 15-second H3 videos per 24 hours. The average paid subscriber does not encounter daily throttling, but concurrency is not the only control: paid plans also use adaptive daily fair-use throughput limits to keep the shared network responsive. Exact thresholds are unpublished to prevent gaming and can shift with GPU supply and demand. Within each plan tier, daily usage is a scheduling signal so accounts with lighter use are favored and more subscribers receive fast service with minimal queue time.

The platform is designed for an individual creator to keep a substantial queue moving throughout the day. Unlimited can queue up to 8 videos at once and Unlimited Pro up to 24; the API and Creative Agent Skill make it practical to automate a personal production queue. Fair use still applies: unattended continuous 24/7 infrastructure and multi-user production workloads require pay-as-you-go Premium Spark or an Enterprise arrangement.

#JavaScript SDK example

Use the exact Turbo workflow ID. Sampling settings are owned by the workflow, so omit steps, guidance, sampler, and scheduler overrides:

const project = await sogni.projects.create({
  type: 'video',
  network: 'fast',
  modelId: 'minimax-h3-fl2va-fp8_t2v_turbo',
  positivePrompt: `integrated_multimodal_description: [Shot 1] A cinematic tracking shot moves through a rainy night market as vendors serve customers beneath glowing awnings.

overall_soundscape: Steady rain, footsteps through shallow puddles, and layered crowd ambience.

non_diegetic_music: A restrained electronic pulse builds beneath the scene.`,
  duration: 8,
  generateAudio: true,
  tokenType: 'spark'
});

const urls = await project.waitForCompletion();

For first-frame animation, use minimax-h3-fl2va-fp8_i2v_turbo with referenceImage. For a transition between two anchors, use minimax-h3-fl2va-fp8_flf2v_turbo with referenceImage and referenceImageEnd.

#Creative Agent Skill examples

The Creative Agent Skill exposes friendly selectors for every current workflow. The generic minimax-h3 and minimax-h3-turbo selectors infer text-to-video, image-to-video, or first/last-frame mode from the supplied frames; standard Ref2VA remains explicit.

# Standard quality-first text-to-video
sogni-agent --video -m minimax-h3 --duration 10 "<three-field H3 prompt>"

# 4-step Turbo image-to-video
sogni-agent --video -m minimax-h3-i2v-turbo --ref first.png --duration 8 "<I2V preamble plus three-field H3 prompt>"

# 4-step Turbo first-and-last-frame video
sogni-agent --video -m minimax-h3-flf2v-turbo --ref first.png --ref-end last.png --duration 8 "<FLF2V preamble plus three-field H3 prompt>"

# Standard Ref2VA with labelled image, video, and audio references
sogni-agent --video -m minimax-h3-r2v --ref identity.png --ref-video motion.mp4 --ref-audio voice.m4a "<six-field Ref2VA prompt>"

Use minimax-h3-turbo when you want the Skill to infer Turbo t2v/i2v/flf2v automatically. There is no Turbo Ref2VA selector.

#Worker hardware

MiniMax H3 and H3 Turbo jobs require a supported NVIDIA GPU with at least 32 GB of VRAM. The RTX 5090 is the entry consumer-class example. Sogni Comfy Worker applies Blackwell-specific runtime tuning to the RTX 5090 and RTX PRO 6000 Blackwell, but H3 is not restricted to Blackwell: other supported NVIDIA GPUs with 32 GB or more VRAM may also be eligible. See the Fast Worker requirements.

MiniMax's hosted H3-Context-IR model stage and 2K regeneration features are not part of the current release: MiniMax has not yet published them as open weights, so current Sogni output should not be described as native 2K. MiniMax has committed to releasing its second-phase internal upscaler for 2K output, and Sogni plans to integrate it as soon as those weights become available. That update will be included automatically for existing subscribers. Unless a release is specifically marked otherwise, newly supported open-weight models and related updates are automatically included in both existing and new subscriptions. The separate H3 Ref2VA reference-to-video checkpoint is already included and runs on the same 32 GB worker class. The hosted model stage is distinct from the ordered Context-IR prompt document used by the released workflows.

#Prompting

Base and Turbo text-to-video, image-to-video, and first/last-frame workflows use MiniMax's ordered three-field prompt contract:

integrated_multimodal_description: [Shot 1] ...

overall_soundscape: ...

non_diegetic_music: ...

Image-to-video and first/last-frame prompts prepend their mode-specific alignment instruction, followed by a blank line before the three fields. Use [Shot N] notation, keep stable (S1), (S2), and subsequent speaker IDs across shots, and write exact dialogue as <d>[Language] words</d>. Put negative direction inside the structured prompt because H3 has no negative-prompt field.

Ref2VA uses a separate six-field contract in this order: subject_definitions, summary, retention_analysis, detailed_description, overall_soundscape, and non_diegetic_music. Its references are loose references rather than locked frames. Address them with H3's 1-based per-type labels — <Picture 1>, <Video 1>, and <Audio 1> — and give each one an explicit job such as identity, wardrobe, environment, camera movement, or voice. State which reference wins when two disagree. A reference video's own soundtrack takes the next <Audio n> number before standalone audio clips.

Official sources: MiniMax H3 model card, MiniMax H3 license, LightX2V MiniMax H3 Turbo model card, LightX2V Turbo implementation, and ComfyUI day-zero H3 support.