MiniCPM-V 4.7 1B is the lighter one. Qwen3 8B needs more memory.
First modelAya Expanse 32B Aya Expanse 8B bge-m3 Chatterbox DeepSeek R1 Distill 32B Distil-Whisper large-v3 FLUX.1 Dev FLUX.1 Schnell FLUX.2 Dev FLUX.2 Klein 4B Gemma 3 12B Gemma 3 27B Gemma 3 4B Gemma 3n E4B Gemma 4 26B A4B Gemma 4 31B gpt-oss 120B gpt-oss 20B Hermes 3 8B Kokoro 82M Llama 3.2 3B Llama 3.3 70B LTX-2.3 MiniCPM-V 4.7 1B MiniCPM-V 4.7 35B-A3B NLLB-200 1.3B nomic-embed-text olmOCR 2 7B PaddleOCR Phi-4 14B Piper (English) Qwen2.5-Coder 32B Qwen2.5-Coder 7B Qwen2.5-VL 7B Qwen3 235B-A22B Qwen3 8B Qwen3-Coder 30B-A3B Qwen3.5 27B Qwen3.5 35B-A3B Qwen3.8 27B Real-ESRGAN Stable Diffusion XL Tesseract Wan 2.1 1.3B Wan 2.1 14B Whisper large-v3 Whisper large-v3 turbo Whisper small Second modelAya Expanse 32B Aya Expanse 8B bge-m3 Chatterbox DeepSeek R1 Distill 32B Distil-Whisper large-v3 FLUX.1 Dev FLUX.1 Schnell FLUX.2 Dev FLUX.2 Klein 4B Gemma 3 12B Gemma 3 27B Gemma 3 4B Gemma 3n E4B Gemma 4 26B A4B Gemma 4 31B gpt-oss 120B gpt-oss 20B Hermes 3 8B Kokoro 82M Llama 3.2 3B Llama 3.3 70B LTX-2.3 MiniCPM-V 4.7 1B MiniCPM-V 4.7 35B-A3B NLLB-200 1.3B nomic-embed-text olmOCR 2 7B PaddleOCR Phi-4 14B Piper (English) Qwen2.5-Coder 32B Qwen2.5-Coder 7B Qwen2.5-VL 7B Qwen3 235B-A22B Qwen3 8B Qwen3-Coder 30B-A3B Qwen3.5 27B Qwen3.5 35B-A3B Qwen3.8 27B Real-ESRGAN Stable Diffusion XL Tesseract Wan 2.1 1.3B Wan 2.1 14B Whisper large-v3 Whisper large-v3 turbo Whisper small
OpenBMB
The small MiniCPM-V 4.7. It looks at pictures, several pictures, a document, or a video. The Q4_K_M language file is about 0.6 GB, and the vision projector brings the running size to about 1.7 GB. The 3 GB phone figure belongs to the 35B-A3B size.
File 0.6 GB
License Apache 2.0
Used for Reading a scan
Quality note OpenBMB's small vision model for pictures, documents, and video. It is not a mixture of experts. Fits on: Windows laptop, 4 GB, Old office PC, 4 GB, MacBook Air, 8 GB, and 26 more
Alibaba
The default small chatbot for 2026. Quick, capable, and easy to find.
File 5.2 GB
License Apache 2.0
Used for Chat and writing, Document search, Translation, Coding
Quality note A step behind Gemma 3 12B, and a clear step ahead of a 3B toy model. Fits on: Windows laptop, 16 GB, Gaming PC, 8 GB GPU, MacBook Air, 8 GB, and 19 more