WeMM-Video-35BA3B: Hardware necesario y cómo ejecutarlo gratis
WeMM-Video-35B-A3B es un modelo de código abierto chino‑inglés para comprensión conjunta de video, audio, imagen y texto. Conoce sus capacidades, rendimiento y…
Qué dice el vídeo
Tencent claims this model is the new open-source state of the art for audio-visual understanding. It can watch a video, listen to its soundtrack, and answer in Chinese or English.
WeMM-Video-35B-A3B is a mixture-of-experts checkpoint that totals 35 billion parameters, with three billion experts active for each token, according to the model card. It belongs to the WeMM-Video family, designed for joint audio-visual understanding and reasoning. The model accepts four modalities: plain text, still images, audio clips, and video files with or without sound. After processing the media, it produces a textual answer in either Chinese or English. The weights are stored as BF16 safetensors and occupy 67.39 GiB. A multilingual approach lets it serve creators who need both languages, and its token-based architecture lets it reason across time-stamped video frames and their accompanying audio.
The new aspect of this model is its mixture-of-experts architecture, which keeps only a few experts active per token, reducing compute while preserving capacity. It uniquely combines video frames with the original soundtrack, allowing it to describe events, transcribe speech, and reason about what is heard and seen together. It also works with silent video, treating visual information alone, and can process standalone audio files. Responses are generated in either Chinese or English, making the model truly bilingual. This flexibility sets it apart from earlier multimodal models that handled only images or text, and it opens up richer interactive applications that need both visual and auditory context.
According to the model card, WeMM-Video-35B-A3B establishes a new open-source state of the art for multimodal tasks. It outperforms proprietary models such as Qwen3.5-Omni-Flash and Seed2.0-Mini-0428, and it narrows the performance gap with Gemini-3.0-Flash. These claims come from the evaluation section, where the authors compare the model against several leading competitors. While exact scores are not listed, the statement positions the model at the top of publicly available video-audio reasoning systems. This makes it a compelling free alternative for anyone needing high-quality multimodal inference without paying for a commercial API.
The model’s biggest strength is its ability to jointly understand and reason over video, audio, images, and text, delivering coherent textual answers in two languages. Performance tests show it can generate about 154 tokens per second on an RTX 4090 and 205 tokens per second on an RTX 5090, when the full weights reside in GPU memory. At 4-bit precision with an 8 K context, it needs roughly 25 GB of memory, which fits on high-end consumer cards. However, the weight files total 67.39 GiB, so loading the full model demands substantial storage and GPU capacity. Also, the model only produces text responses, so it cannot directly generate images or audio as output.
If you run AI experiments at home and have access to a powerful GPU, this model is a solid choice. Its bilingual capability makes it useful for creators producing content in Chinese and English. Researchers looking for a free, open-source video-audio reasoning system will appreciate the state-of-the-art claims without paying for a proprietary API. The model runs best on machines like an RTX 4090 or RTX 5090, where token throughput is highest, but it also works at lower speeds on older cards if you accept slower generation. Anyone interested in exploring multimodal AI without licensing fees should give WeMM-Video-35B-A3B a try.
¿Qué necesitas para ejecutarlo en casa? A 4 bits ocupa unos 25 gigabytes de memoria. El equipo doméstico más pequeño que lo aloja por completo es un RTX 5090. Con un RTX 3060 12 GB y 32 gigabytes de RAM escribe unos 39 tokens por segundo. Con un RTX 4090 y 64 gigabytes de RAM escribe unos 154 tokens por segundo.
Aún no hay API gratuita para ello. Los pesos son abiertos: puedes descargarlos de Hugging Face bajo la licencia Apache 2.0. Arena aún no lo ha clasificado: añadiremos la puntuación en cuanto lleguen los votos.
Compare WeMM Video 35BA3B with every free model, and check what your own GPU can run, on FreeLLMArena.
Los datos
- Tamaño: 36.2B parámetros (4.5B activos) · contexto 256K · memoria a 4-bit alrededor de 25 GB.
- RTX 3060 12 GB + 32 GB de RAM: unos 39 tokens/s
- RTX 4090 + 64 GB de RAM: unos 154 tokens/s
- RTX 5090 + 64 GB de RAM: unos 205 tokens/s
- Mac M4 Max 128 GB: unos 75 tokens/s
- NVIDIA DGX Spark 128 GB: unos 61 tokens/s
- Ryzen AI Max+ 395 128 GB: unos 58 tokens/s
- Mac Studio M3 Ultra 512 GB: unos 91 tokens/s
- Ficha del modelo: https://huggingface.co/tencent/WeMM-Video-35BA3B
- Todos los modelos gratis, comparados: https://freellmarena.com
Hardware
Memoria a 4 bits con un contexto de 8K, de nuestro simulador. Para comprobar tu propio equipo, usa el simulador Ejecútalo en casa.
- WeMM-Video-35BA3B: unos 25 GB a 4 bits · el equipo doméstico más pequeño que lo aloja: RTX 5090 (32 GB) · Amazon ↗
- RTX 3060 12 GB + 32 GB RAM: WeMM-Video-35BA3B, alrededor de 39 tokens/s (GPU + RAM del sistema) · Amazon ↗
- RTX 4090 (24 GB) + 64 GB de RAM: WeMM-Video-35BA3B, unos 154 tokens/s (todo en la GPU) · Amazon ↗
- RTX 5090 (32 GB) + 64 GB de RAM: WeMM-Video-35BA3B, unos 205 tokens/s (todo en la GPU) · Amazon ↗
- Mac M4 Max 128 GB: WeMM-Video-35BA3B, unos 75 tokens/s (todo en la GPU) · Amazon ↗
- NVIDIA DGX Spark 128 GB: WeMM-Video-35BA3B, unos 61 tokens/s (todo en la GPU) · Amazon ↗
- Ryzen AI Max+ 395 128 GB: WeMM-Video-35BA3B, unos 58 tokens/s (todo en la GPU) · Amazon ↗
- Mac Studio M3 Ultra 512 GB: WeMM-Video-35BA3B, unos 91 tokens/s (todo en la GPU) · Amazon ↗
Como afiliados de Amazon, FreeLLMArena gana con las compras que cumplen los requisitos; los enlaces no cambian las cifras.
Every number in this post comes from the FreeLLMArena data on 20 Nov 2026: Arena votes, Hugging Face sizes and our speed simulator. Download the video and its subtitles.
Todos los modelos de IA gratis, clasificados
Tamaños, memoria, velocidad en tu tarjeta gráfica y dónde usar cada modelo gratis. Se actualiza cada noche.