# WeMM-Video-2B

> WeMM-Video-2B is the dense 2B checkpoint in the WeMM-Video family. WeMM-Video is a Chinese–English bilingual family for joint audio-visual understanding and reasoning. It accepts text, images, audio, and video and generates text responses.

- **Organisation:** tencent
- **Category:** Vision-Language Models
- **Parameters:** 3.3B
- **Context length:** 262,144 tokens
- **Licence:** Apache 2.0 (Commercial use OK)
- **Released:** 2026-11-17
- **Hugging Face:** 259 downloads (30 days), 10 likes

## Estimated memory

| Quant | ≈ GB |
|---|---|
| Q2_K | 1.93 |
| Q3_K_M | 2.44 |
| Q4_K_M | 2.82 |
| Q5_K_M | 3.16 |
| Q6_K | 3.52 |
| Q8_0 | 4.29 |
| F16 | 7.3 |

## Runs with

- [llama.cpp](https://selfhost.directory/project/llamacpp)
- [LM Studio](https://lmstudio.ai)
- [Transformers](https://github.com/huggingface/transformers)
- [Jan](https://jan.ai) (community)
- [KoboldCpp](https://selfhost.directory/project/koboldcpp) (community)
- [LocalAI](https://selfhost.directory/project/localai) (community)
- [Ollama](https://selfhost.directory/project/ollama) (community)

## About

WeMM-Video-2B is the dense 2B checkpoint in the WeMM-Video family. WeMM-Video is a Chinese–English bilingual family for joint audio-visual understanding and reasoning. It accepts text, images, audio, and video and generates text responses.

• Weights · BF16 Safetensors · 6.12 GiB of weight files
• Quantized variants · FP8 Block128 · AWQ W4A16 · GGUF Q4_K_M · NVFP4 W4A16

WeMM-Video can combine video frames with the original soundtrack to describe events, transcribe speech, and reason about what is seen and heard. It also accepts silent video, standalone audio, images, and text, with responses in Chinese or English.

Vision and audio encoders turn media into tokens that the language model processes alongside text. Video frames carry timestamps. The family was pretrained on approximately 1.2 trillion multimodal tokens, then post-trained with multi-teacher OPD, anti-repetition MrDPO, and specialist reinforcement learning.

See the project overview for family details.

Set up the Transformers environment. From the WeMM-Video repository root, run the example on an MP4 with a soundtrack. The script loads the model and processor from the Hub and processes both the frames and the original audio track.

Use --revision to pin the model and processor to a Hub commit. Size GPU memory for the full weight files, media inputs, and generation cache.

## Alternatives

- [Qwen/Qwen3.5-2B](https://selfhost.directory/models/qwen/qwen3.5-2b) (2.3B, Apache 2.0, ≈2.22 GB at Q4): Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
- [Qwen/Qwen3.5-4B](https://selfhost.directory/models/qwen/qwen3.5-4b) (4.7B, Apache 2.0, ≈4.26 GB at Q4): Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
- [LiquidAI/LFM2.5-VL-3B](https://selfhost.directory/models/liquidai/lfm2.5-vl-3b) (3.1B, lfm1.0, ≈2.82 GB at Q4): LFM2.5-VL-3B is a multimodal variant of LFM2.5, a family of hybrid models designed for on-device deployment. It builds on LFM2-VL-3B with further mid- and post-training.
- [LiquidAI/d1-3B](https://selfhost.directory/models/liquidai/d1-3b) (3.1B, lfm1.0, ≈2.82 GB at Q4): d1-3B is a 3B parameter decision model built on LFM2.5-VL-3B. You give it a state (text, JSON, images, or a mix) and a set of questions. It returns calibrated, typed answers in one forward pass with zero output tokens.
- [Qwen/Qwen3-VL-4B-Instruct](https://selfhost.directory/models/qwen/qwen3-vl-4b-instruct) (4.4B, Apache 2.0, ≈4.26 GB at Q4): Meet Qwen3-VL — the most powerful vision-language model in the Qwen series to date.
- [meta-models/Muse-Glimmer-30B-assistant](https://selfhost.directory/models/meta-models/muse-glimmer-30b-assistant) (2.6B, Apache 2.0, ≈2.17 GB at Q4): Muse Glimmer is a 30-billion-parameter causal language model with a dedicated perception encoder, distilled from Muse Spark and purpose-built for autonomous agentic tasks on consumer hardware.

More: https://selfhost.directory/models/tencent/wemm-video-2b/alternatives

## Runs well on

- [GeForce RTX 4060](https://selfhost.directory/models/tencent/wemm-video-2b/on/rtx-4060): F16, est. 23 tok/s
- [Apple M4 (16 GB)](https://selfhost.directory/models/tencent/wemm-video-2b/on/apple-m4-16gb): F16, est. 10 tok/s
- [Radeon RX 6700 XT](https://selfhost.directory/models/tencent/wemm-video-2b/on/rx-6700-xt): F16, est. 32 tok/s
- [GeForce RTX 3060 12GB](https://selfhost.directory/models/tencent/wemm-video-2b/on/rtx-3060): F16, est. 30 tok/s
- [Arc B580](https://selfhost.directory/models/tencent/wemm-video-2b/on/arc-b580): F16, est. 38 tok/s
- [Tesla T4](https://selfhost.directory/models/tencent/wemm-video-2b/on/tesla-t4): F16, est. 27 tok/s

## Similar size

- [Qwen/Qwen2.5-VL-3B-Instruct](https://selfhost.directory/models/qwen/qwen2.5-vl-3b-instruct) (3.8B)
- [nvidia/LocateAnything-3B](https://selfhost.directory/models/nvidia/locateanything-3b) (3.8B)
- [LiquidAI/d1-3B](https://selfhost.directory/models/liquidai/d1-3b) (3.1B)
- [nvidia/Cosmos-Reason2-2B](https://selfhost.directory/models/nvidia/cosmos-reason2-2b) (2.4B)
- [LiquidAI/LFM2.5-VL-3B](https://selfhost.directory/models/liquidai/lfm2.5-vl-3b) (3.1B)
- [ibm-granite/granite-vision-3.3-2b](https://selfhost.directory/models/ibm-granite/granite-vision-3.3-2b) (3B)

## FAQ

**How much VRAM does WeMM-Video-2B need?**
About 2.8 GB at Q4_K_M with an 8K context. Q8_0 needs about 4.3 GB and full precision about 7.3 GB. Longer contexts need more memory for the KV cache.

**Can WeMM-Video-2B run on the RTX 4060?**
Yes. It fits at FP16 / BF16 (about 7.3 GB), at an estimated 23 tokens/s. It fits comfortably on the Apple M4 (16 GB), Radeon RX 6700 XT, RTX 3060 12GB.

**Can I use WeMM-Video-2B commercially?**
Yes. It is released under Apache 2.0. Permissive: use, modify and sell freely; keep the notice and licence text.

**What is the context length of WeMM-Video-2B?**
262,144 tokens (256K). The full window needs far more memory than an 8K context.

## Links

- huggingface: https://huggingface.co/tencent/WeMM-Video-2B
- github: https://github.com/Tencent/WeMM-Video
- paper: https://arxiv.org/abs/2611.26409

---
Source: https://selfhost.directory/models/tencent/wemm-video-2b