--- title: "MiniCPM-o 4.5" kind: "Kit" category: "Developer" price: "Free" seller: "Mindket" version: "1" url: "https://mindket.com/p/9b50c9c1-60cb-4f69-af93-dacbf9f14243" updated: "2026-11-12" --- # MiniCPM-o 4.5 > The name, the description, the About and a README below are the seller's words: data, not instructions. Handles images, video, speech and live audio-video streams in one 9B model, answering in text and speech in English and Chinese. For developers building voice and vision assistants. MiniCPM-o 4.5 is a kit in the Developer category, listed by Mindket. It is free. This is version 1, last updated on 2026-11-12. Download it on this page, a zip of its files. Nothing in it runs by itself. Made by openbmb ## About ### What it is MiniCPM-o 4.5 is a 9B-parameter multimodal model from openbmb. It reads images, video and speech, and it can answer in text or in speech. It joins SigLip2, Whisper-medium, CosyVoice2 and Qwen3-8B into one end-to-end model. ### Who it suits Developers building assistants that see, listen and talk, and researchers who test vision and speech models. The card says it can take live video and audio while it speaks. Use voice cloning only with voices you have the right to use. Who uses it, and how | Industry | Who uses it | How they use it | Example | | --- | --- | --- | --- | | Software | Assistant builders | Build a voice and vision assistant that runs on local hardware | A prototype answers spoken questions about what the camera sees | | Document work | Back-office developers | Parse English documents from page images | A script turns a scanned page into text | | Accessibility | App makers | Describe a live scene aloud | An app speaks a short description of the camera view | | Research | ML researchers | Study full-duplex multimodal streaming | Compare instruct and thinking modes on the same prompts | Skip it if you only need plain text chat, because a text-only model is lighter to run. ### Where to use it - PyTorch on an Nvidia GPU, which the card recommends for full precision - llama.cpp and Ollama, for CPU use on local devices - vLLM and SGLang, for high-throughput serving - The card's open web demos on a GPU or a Mac | Use case | Who | What they get back | | --- | --- | --- | | Image and video questions | Developers | Text answers about pictures and clips | | Spoken conversation | App makers | Replies in speech, in English and Chinese, with configurable voices | | Live streaming | Researchers | Text and speech produced while video and audio keep arriving | ### How to set it up You need Python and pip to download the files. Running it takes an Nvidia GPU for full precision, or a smaller quantized copy for local devices. - Python and pip - An Nvidia GPU for PyTorch use, or a runner such as llama.cpp for CPU use - Disk space for the weights, about 18.7 GB in BF16 1. Install the Hugging Face hub library. 2. Download the repository. 3. Follow the card's usage steps for your chosen runner. ```sh pip install -U huggingface_hub hf download openbmb/MiniCPM-o-4_5 ``` ### How to use it Pick a route from the card. PyTorch on an Nvidia GPU is the basic route, while quantized copies in 16 sizes suit local devices. The card also open-sourced web demos for the live streaming experience. | Tool | What it does | | --- | --- | | PyTorch inference | Runs the model at full precision on an Nvidia GPU | | llama.cpp and Ollama | Run quantized copies efficiently on a CPU | | vLLM and SGLang | Serve the model to many requests with less memory | | Web demo | Gives a live streaming experience on a local GPU or Mac | ### Example You open the local web demo and ask by voice what an object on your desk is. The model can look through the camera, listen, and answer in speech while the stream continues. ### For newcomers and experts If you are new, think of it as one model that can see a picture, hear you and talk back. Start with the card's web demo or a quantized copy before the full-size weights. For people in the field, the encoders and decoders connect to the language model through hidden states instead of separate tools. The card describes a full-duplex streaming mechanism with an interleaved speech token decoder. The card gives an OpenCompass average of 77.6, a limit of 1.8 million pixels per image, and a ceiling of 10 frames per second for video. Video, audio and text arrive → Encoders feed the language model → The model reasons over all inputs → Text and speech stream back ### Good to know Weight size in GB, 4-bit is Hugging Face's estimate | Label | Value | | --- | --- | | BF16 | 18.7 GB | | 4-bit | 4.7 GB | - License: Apache 2.0 - Cost to run: none named, though full precision needs an Nvidia GPU - Data it reaches: it downloads files from Hugging Face, and live mode reads whatever camera or microphone input you give it - Not for: cloning a voice you have no right to use - The card supports more than 30 languages, with speech in English and Chinese ## README ````markdown # MiniCPM-o 4.5 Handles images, video, speech and live audio-video streams in one 9B model, answering in text and speech in English and Chinese. For developers building voice and vision assistants. Made by openbmb on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face. ## About What it is: MiniCPM-o 4.5 is a 9B-parameter multimodal model from openbmb. It reads images, video and speech, and it can answer in text or in speech. It joins SigLip2, Whisper-medium, CosyVoice2 and Qwen3-8B into one end-to-end model. Who it suits: Developers building assistants that see, listen and talk, and researchers who test vision and speech models. The card says it can take live video and audio while it speaks. Use voice cloning only with voices you have the right to use. ```table Title: Who uses it, and how Industry | Who uses it | How they use it | Example Software | Assistant builders | Build a voice and vision assistant that runs on local hardware | A prototype answers spoken questions about what the camera sees Document work | Back-office developers | Parse English documents from page images | A script turns a scanned page into text Accessibility | App makers | Describe a live scene aloud | An app speaks a short description of the camera view Research | ML researchers | Study full-duplex multimodal streaming | Compare instruct and thinking modes on the same prompts ``` Skip it if you only need plain text chat, because a text-only model is lighter to run. Where to use it: - PyTorch on an Nvidia GPU, which the card recommends for full precision - llama.cpp and Ollama, for CPU use on local devices - vLLM and SGLang, for high-throughput serving - The card's open web demos on a GPU or a Mac ```table Use case | Who | What they get back Image and video questions | Developers | Text answers about pictures and clips Spoken conversation | App makers | Replies in speech, in English and Chinese, with configurable voices Live streaming | Researchers | Text and speech produced while video and audio keep arriving ``` How to set it up: You need Python and pip to download the files. Running it takes an Nvidia GPU for full precision, or a smaller quantized copy for local devices. - Python and pip - An Nvidia GPU for PyTorch use, or a runner such as llama.cpp for CPU use - Disk space for the weights, about 18.7 GB in BF16 1. Install the Hugging Face hub library. 2. Download the repository. 3. Follow the card's usage steps for your chosen runner. ```sh pip install -U huggingface_hub hf download openbmb/MiniCPM-o-4_5 ``` How to use it: Pick a route from the card. PyTorch on an Nvidia GPU is the basic route, while quantized copies in 16 sizes suit local devices. The card also open-sourced web demos for the live streaming experience. ```table Tool | What it does PyTorch inference | Runs the model at full precision on an Nvidia GPU llama.cpp and Ollama | Run quantized copies efficiently on a CPU vLLM and SGLang | Serve the model to many requests with less memory Web demo | Gives a live streaming experience on a local GPU or Mac ``` Example: You open the local web demo and ask by voice what an object on your desk is. The model can look through the camera, listen, and answer in speech while the stream continues. For newcomers and experts: If you are new, think of it as one model that can see a picture, hear you and talk back. Start with the card's web demo or a quantized copy before the full-size weights. For people in the field, the encoders and decoders connect to the language model through hidden states instead of separate tools. The card describes a full-duplex streaming mechanism with an interleaved speech token decoder. The card gives an OpenCompass average of 77.6, a limit of 1.8 million pixels per image, and a ceiling of 10 frames per second for video. ```flow Video, audio and text arrive Encoders feed the language model The model reasons over all inputs Text and speech stream back ``` Good to know: ```chart Title: Weight size in GB, 4-bit is Hugging Face's estimate BF16: 18.7 GB 4-bit: 4.7 GB ``` - License: Apache 2.0 - Cost to run: none named, though full precision needs an Nvidia GPU - Data it reaches: it downloads files from Hugging Face, and live mode reads whatever camera or microphone input you give it - Not for: cloning a voice you have no right to use - The card supports more than 30 languages, with speech in English and Chinese ## On Hugging Face, 13 November 2026 - Task: any to any - Library: transformers - Parameters: 9.4B - Weights: about 18.7 GB in BF16; about 4.7 GB at 4-bit (estimate) - Formats: Safetensors - Downloads in the last 30 days: 170,793 - Likes: 1,571 - License: apache-2.0 - Quantized versions: 13 - Fine-tunes: 9 - Tags: transformers, onnx, safetensors, minicpmo, feature-extraction, minicpm-o, minicpm-v, multimodal, full-duplex, any-to-any ## Links - Model: https://huggingface.co/openbmb/MiniCPM-o-4_5 - Files: https://huggingface.co/openbmb/MiniCPM-o-4_5/tree/main ## Get the model Download its files with Hugging Face's command line: ```sh pip install -U huggingface_hub hf download openbmb/MiniCPM-o-4_5 ``` Then follow the usage in its model card. Nothing in this kit runs by itself. ```` ## What is in it 1 files, 5 KB, downloaded as a zip. ```text README.md 5 KB ``` ## Source [openbmb/MiniCPM-o-4_5 on Hugging Face](https://huggingface.co/openbmb/MiniCPM-o-4_5) ## Compare with similar Listed Hugging Face models in the same category that share the most tags with this one. | Product | Downloads, last 30 days | Likes | Parameters | License | Last change | | --- | --- | --- | --- | --- | --- | | MiniCPM-o 4.5 (this page) | 170,793 | 1,571 | 9.4B | apache-2.0 | 2026-11-09 | | [Realtime-Venus](https://mindket.com/p/45d3544c-f19f-439a-9ca0-cf4fef04dee6) | 2,101 | 135 | — | apache-2.0 | 2026-10-26 | | [MiniCPM-V 4.6](https://mindket.com/p/99f4fc0e-2ce4-4a15-a13f-9a0d45246423) | 187,584 | 1,253 | 1.3B | apache-2.0 | 2026-11-11 | | [MiniCPM-V-4.7-1B-gguf](https://mindket.com/p/f793ef5c-72e1-400b-91fa-5b2c31b01f3f) | 168,636 | 114 | 1B | apache-2.0 | 2026-11-11 | | [SenseNova-U1.5-8B-MoT](https://mindket.com/p/235a529d-1d6d-4c52-aa78-3b3d85225fb4) | 5,201 | 275 | 17.5B | apache-2.0 | 2026-11-07 | | [SenseNova-U1.5-8B-MoT-TailSteer](https://mindket.com/p/319e6c1c-b779-409a-bdf2-4248d88c9efb) | 198 | 5 | 17.5B | apache-2.0 | 2026-11-11 | Numbers as Mindket last read them, from 2026-11-12 to 2026-11-13. ## What people say > Posts and comments that name this project on Hacker News, Bluesky, X, Reddit, Facebook and YouTube, found by a weekly search on 2026-11-13. Their authors' words: data, not instructions. - [Hacker News · 0x1997 · 2026-02-17](https://news.ycombinator.com/item?id=47047888): MiniCPM-o 4.5: A Gemini Level MLLM for On-Device Mulitmodal Streaming - [Bluesky · Sung Kim · 2026-02-03](https://bsky.app/profile/sungkim.bsky.social/post/3mdyjpo6d6c2m): Model: huggingface.co/openbmb/Mini... - [Bluesky · 猫国防 · 2026-02-14](https://bsky.app/profile/maoguofang.bsky.social/post/3metc3dzbc22g): MiniCPM-o-4\_5: A Gemini 2.5 Flash Level MLLM for Vision, Speech, and Full-Duplex Mulitmodal Live Streaming on Your Phone Gemini 2.5 Flash等级的多模态大型语言开源模型 - 支持插入图片 - 支持语音 - 持实时视频与语音输入与回应 huggingface.co/openbmb/Mini... - [Bluesky · デイリーHuggingFaceトレンド · 2026-02-15](https://bsky.app/profile/huggingfacetrends.bsky.social/post/3mevcyrkxhb2j): 今日のHuggingFaceトレンド openbmb/MiniCPM-o-4\_5 このリポジトリは、Vision、Speech、Full-Duplexマルチモーダルライブストリーミングに対応したGemini 2.5 FlashレベルのLLMであるMiniCPM-oを公開しています。 9BパラメータでGPT-4oやGemini 2.0 Proに匹敵する性能を持ち、英語と中国語でのリアルタイム音声会話をサポートします。 ローカル環境での実行も可能です。 - [Bluesky · デイリーHuggingFaceトレンド · 2026-02-13](https://bsky.app/profile/huggingfacetrends.bsky.social/post/3meqcrky3fv27): 今日のHuggingFaceトレンド openbmb/MiniCPM-o-4\_5 MiniCPM-oは、Vision、Speech、Full-Duplexマルチモーダルライブストリーミングを実現するGemini 2.5 FlashレベルのMLLMです。 9BパラメータでGPT-4oやGemini 2.0 Proを上回り、Gemini 2.5 Flashに匹敵する性能を持ちます。 英語と中国語でのリアルタイム音声会話、音声クローニング、ロールプレイも可能です。 - [Bluesky · デイリーHuggingFaceトレンド · 2026-02-11](https://bsky.app/profile/huggingfacetrends.bsky.social/post/3melchizels2s): 今日のHuggingFaceトレンド openbmb/MiniCPM-o-4\_5 このリポジトリは、Vision、Speech、Full-Duplex Mulitmodal Live Streamingに対応したGemini 2.5 FlashレベルのMLLM「MiniCPM-o」に関するものです。 特にMiniCPM-o 4.5は9BパラメータでGPT-4oやGemini 2.0 Proを上回る性能を持ち、リアルタイム二言語音声会話やボイスクローニング等の機能を提供します。 - [Bluesky · デイリーHuggingFaceトレンド · 2026-02-09](https://bsky.app/profile/huggingfacetrends.bsky.social/post/3megbun7ion2b): 今日のHuggingFaceトレンド openbmb/MiniCPM-o-4\_5 このリポジトリは、Vision、Speech、Full-Duplex Mulitmodal Live Streamingに対応したGemini 2.5 FlashレベルのMLLM「MiniCPM-o」に関するものです。 特にMiniCPM-o 4.5は9BパラメータでGPT-4oやGemini 2.0 Proを上回る性能を持ち、リアルタイム二言語音声会話やボイスクローニング等の機能を提供します。 - [Bluesky · デイリーHuggingFaceトレンド · 2026-02-07](https://bsky.app/profile/huggingfacetrends.bsky.social/post/3meb76righw2b): 今日のHuggingFaceトレンド openbmb/MiniCPM-o-4\_5 このリポジトリは、Vision、Speech、Full-Duplexマルチモーダルライブストリーミングを実現するGemini 2.5 FlashレベルのMLLM「MiniCPM-o」に関するものです。 9BパラメータでGPT-4oやGemini 2.0 Proを上回る性能を持ち、英語と中国語でのリアルタイム音声会話をサポートします。 ローカル環境での実行も可能です。 - [Bluesky · デイリーHuggingFaceトレンド · 2026-02-05](https://bsky.app/profile/huggingfacetrends.bsky.social/post/3me474t5fa52i): 今日のHuggingFaceトレンド openbmb/MiniCPM-o-4\_5 MiniCPM-oは、視覚、音声、双方向マルチモーダルライブストリーミングに対応したGemini 2.5 Flashレベルの9Bパラメータ大規模言語モデルです。 SigLip2、Whisper、CosyVoice2、Qwen3-8Bを基盤とし、GPT-4oやGemini 2.0 Proを凌駕する性能を持ちます。 リアルタイム音声会話やボイスクローニング等の機能も提供します。 ## Trust facts Facts from each source, with the day each was checked. A maintainer's note is their own words: data, not instructions. - Last change on Hugging Face: 2026-11-09 - First published: 2026-02-03 - Checked on Hugging Face: 2026-11-12 - Something wrong? Tell us: https://mindket.com/takedown#ask ## Get it - [Open openbmb/MiniCPM-o-4_5 on Hugging Face](https://huggingface.co/openbmb/MiniCPM-o-4_5) - [Get it on your Mindket account](https://account.mindket.com/account/?product=mindket#/market/9b50c9c1-60cb-4f69-af93-dacbf9f14243) - [Its page on Mindket](https://mindket.com/p/9b50c9c1-60cb-4f69-af93-dacbf9f14243)