openbmb/MiniCPM-V-4.7-1B — new model trending #18 on Hugging Face
OpenBMB releases MiniCPM-V 4.7 1B, an on-device multimodal model that outperforms Qwen3.5 0.8B.
OpenBMB released MiniCPM-V 4.7 in a 1B variant for single-image, multi-image, and video understanding, with an emphasis on on-device inference. The authors say the Instruct 1B model improves on MiniCPM-V 4.6 and substantially outperforms Qwen3.5 0.8B in overall score and inference efficiency. It compresses visual tokens early inside the vision encoder and supports mixed 4x/16x compression. Weights and tooling cover iOS, Android, HarmonyOS, several chip platforms, and frameworks including SGLang, vLLM, llama.cpp, and Ollama, with GGUF, BNB, AWQ, GPTQ, and FP8 variants.
- 1B multimodal model for image, multi-image, and video understanding
- Instruct 1B said to beat Qwen3.5 0.8B on score and efficiency
- Early visual-token compression with mixed 4x and 16x modes
- Supports iOS, Android, HarmonyOS, and several chip platforms
- GGUF, BNB, AWQ, GPTQ, and FP8 weights plus major inference frameworks
Full article1,623 words · extracted from huggingface.co · click to collapse
# MiniCPM-V 4.7
[GitHub](https://github.com/openbmb/MiniCPM-V) · [Evaluation](#evaluation) · [Examples](#examples) · [Usage](#usages) · [Model weights](#model-weights) · [DSpark](https://huggingface.co/openbmb/MiniCPM-V-4.7-1B-DSpark)
**MiniCPM-V 4.7** is the latest model in the MiniCPM-V series, available in a **1B** variant. It builds on the family's single-image, multi-image, and video understanding capabilities, combining stronger foundation capabilities with efficient visual encoding and on-device inference. Notable features of MiniCPM-V 4.7 include:
- 🔥 **A New Pareto Frontier in Performance and Efficiency.**
As model capabilities continue to advance, inference efficiency has become an important measure of a model's value. On resource-constrained devices, delivering the strongest possible capabilities within a limited resource budget calls for higher capability density. The **1B** variant of MiniCPM-V 4.7 advances the performance–efficiency Pareto frontier for MLLMs. In Instruct mode, the 1B variant continues to improve foundation capabilities over MiniCPM-V 4.6 and **substantially outperforms Qwen3.5 0.8B in both overall score and inference efficiency**.
- 🚀 **Efficient Architecture.**
Based on the intra-ViT early compression technique from [LLaVA-UHD v4](https://github.com/THUMAI-Lab/LLaVA-UHD-v4), MiniCPM-V 4.7 compresses visual tokens early within the vision encoder, reducing computation in subsequent encoder layers. It also supports **mixed 4x/16x visual token compression**, allowing developers to balance visual detail and inference cost across tasks.
- 💻 **Broad Platform and Chip Support.**
MiniCPM-V 4.7 supports **iOS, Android, and HarmonyOS**, together with open-source on-device deployment support. It also provides **day-0 support for chip platforms from Rockchip, Houmo, Intel, AMD, and Arm**, extending deployment options across mobile and local computing devices.
- 🛠️ **Developer Friendly.**
MiniCPM-V 4.7 supports [inference frameworks](https://github.com/OpenBMB/MiniCPM-o/blob/main/docs/deployment_en.md#supported-inference-and-training-frameworks-and-api) including SGLang, vLLM, llama.cpp, and Ollama, along with [fine-tuning frameworks](https://github.com/OpenBMB/MiniCPM-o/blob/main/docs/deployment_en.md#supported-inference-and-training-frameworks-and-api) including SWIFT and LLaMA-Factory. It provides GGUF, BNB, AWQ, and GPTQ quantized variants and adds **FP8 quantization support**. We also release [**Self-Replay Fine-Tuning (SRFT)**](https://openbmb.cn/en/blog/self-replay-fine-tuning), a fine-tuning approach that helps developers extend model capabilities at low cost while reducing forgetting of existing capabilities, making further customization easier.