Primary use stated or inferred cautiously from official documentation.
Duo 1.0 Causal Base
Complete Duo 1.0 bundle for research into audiovisual conversations with simultaneous listening and responding. It combines an intent module with causal audiovisual generation; user audio is transcribed and aligned in the offline implementation.
The essential
Maximum input capacity when published by the source.
Documented maximum generation limit.
Verified availability channels.
Primary functional family and verified specialties.
Declared terms for API access, model weights, or self-hosting.
Limits and integration
| API ID | inclusionAI/Duo-1.0-causal-base |
|---|---|
| Model type | Video · Audio · Multimodal |
| Access model | Downloadable weights with restrictions |
| License | MiniMax H3 Community License · territorial restrictions |
| Deployment | Self-hosted |
| Release | 13 NOV 2026 |
| Knowledge cutoff | Not published |
| Entrance | Video · Audio · Text · Image |
| Exit | Video · Audio |
| Context window | Chunk-based audiovisual interaction; maximum not published |
| maximum output | Synchronized audio and video; maximum duration not published |
| Reasoning | Not published |
| Published tools | |
| Structured Outings | Not published |
| Batch processing | Not published |
| Prompt cache | Not published |
| Fine-tuning | Not published |
| Verified platforms | Hugging Face · vLLM-Omni · SGLang-Omni |
Study participation decisions and synchronized audiovisual responses. Includes intent, adapter, Realization, codecs and speech models. Realization uses PyTorch distributed checkpoints and a 50-step Euler sampler; it is not a production-ready chat API.
The weights license excludes the EU, the UK, South Korea and the United States; use there requires additional authorization. The code is Apache 2.0, but this does not change the terms for the weights or other components. The published version is a conversation preview; features announced for Duo v2 are not yet available. The offline reference requires Linux, Python 3.10, FFmpeg and two CUDA GPUs; a minimum VRAM requirement has not been verified.
What it can do
Orientation
Study participation decisions and synchronized audiovisual responses. Includes intent, adapter, Realization, codecs and speech models. Realization uses PyTorch distributed checkpoints and a 50-step Euler sampler; it is not a production-ready chat API.
Context
Chunk-based audiovisual interaction; maximum not published · Synchronized audio and video; maximum duration not published
Tools and integration
Not published
Access
Hugging Face · Duo implementation
Documented cost
| Concept | Worth | Unit/condition |
|---|---|---|
| Standard input | Not published | Standard API rate |
| Cached input | Not published | Reading reused prefixes |
| Cache write or storage | Not published | The condition varies by provider |
| Standard output | Not published | May include reasoning tokens |
| Batch input | Not published | Asynchronous processing |
| Batch output | Not published | Asynchronous processing |
No hosted API price has been verified. Downloading weights does not remove infrastructure costs or license restrictions. Commercial products exceeding USD 20 million in annual revenue require additional written authorization.
i Prices change and may depend on level, region or context length. Check the source before making a decision.
How to read the results
The report evaluates 30 cases against six criteria: 168 of 180 checks pass. It does not break down results by released checkpoint, so this variant is not assigned an individual score. Inferama has not reproduced the tests.
i Joint human evaluation; no separately verified score per checkpoint.
Inferama only highlights a “best result” when the metric, test set, configuration, and date allow for an equivalent comparison. The supplier's figures are presented as claims from its own source.
Chronology
inclusionAI/Duo-1.0-causal-base
Complete bundle; distilled Duo 1.0 repositories depend on its shared components.