AIGC.NEWSChina AI, from the inside
···
PolicyChina AI LinesEvent History
AI LABS · MAINLAND CHINA · VOICE & SPEECH · LANGUAGE MODELS · MULTIMODAL UNDERSTANDING

ModelBest 面壁智能

Compact language, vision-language and speech models designed for on-device applications. modelbest.cn · github.com

Use & access

Official sitemodelbest.cnHugging Face: openbmbGitHub: OpenBMB

Product links & documentation

Model access and license evidence applies to the named model. Check the official terms for availability in your country.

Products

Model family

  • MiniCPM

    Compact open-weight multimodal model family for on-device use.

    Source

Verified model profiles

Model-specific conditions and evidence

Recent model release coverage

Published Events; a news report does not establish a verified model profile.

  1. Reportedly: ModelBest releases MiniCPM-V 4.7, an on-device multimodal model with 13+ token/s decoding at 3GB memory
  2. OpenBMB released Trident-Qwen3VL-2B, a multimodal embedding model for mixed-modality retrieval
  3. OpenBMB released MiniCPM-V-4.7-35B-A3B-DSpark, a speculative decoding draft model, on Hugging Face
  4. OpenBMB released MiniCPM-V-4.7-1B-DSpark, a speculative decoding draft model for MiniCPM-V-4.7-1B
  5. openbmb released MiniCPM-V-4.7-35B-A3B, the first MoE model in the MiniCPM-V series

Latest Events

  1. OpenBMB releases MiniCPM-OCR 1.0, an open-source 1.3B document parsing model
  2. OpenBMB releases JustGoal GPU kernels outperforming expert libraries on NVIDIA B200
  3. Reportedly: ModelBest releases MiniCPM-V 4.7, an on-device multimodal model with 13+ token/s decoding at 3GB memory
  4. OpenBMB releases llama.cpp-omni, an open-source C++ engine for full-duplex omni-modal streaming
  5. OpenBMB released Trident-Qwen3VL-2B, a multimodal embedding model for mixed-modality retrieval
  6. OpenBMB releases Trident repository for mixed-modality retrieval
  7. OpenBMB released MiniCPM-V-4.7-35B-A3B-DSpark, a speculative decoding draft model, on Hugging Face
  8. OpenBMB released MiniCPM-V-4.7-1B-DSpark, a speculative decoding draft model for MiniCPM-V-4.7-1B
  9. openbmb released MiniCPM-V-4.7-35B-A3B, the first MoE model in the MiniCPM-V series
  10. OpenBMB releases MiniCPM-V-4.7-1B, a 1B-parameter multimodal model for on-device use
  11. OpenBMB open-sources T2Mem for test-time memory construction
  12. ModelBest released JustRL-II-base-model on Hugging Face
  13. ModelBest released UltraData-Code-L2-Classifier on Hugging Face
  14. OpenBMB released MiniCPM5-2B-DSpark, a speculative decoding draft model for MiniCPM5-2B
  15. openbmb launches MiniCPM5-2B, a 2B on-device model with Apache-2.0 weights
  16. ModelBest open-sources Meshy, an asynchronous RL engine for LLMs
  17. OpenBMB released MiniCPM5-2B-Midtrain, a 2B text-generation model for on-device use
  18. OpenBMB launches MiniCPM5-2B-SFT, a 2B open-weights model for on-device text generation
  19. OpenBMB released MathForm-8B, an autoformalization model for Lean 4