MiniCPM O 4 5 — Hardware Requirements & GPU Compatibility
ChatFunctions
MiniCPM-o 4.5 is OpenBMB's roughly 9.4-billion-parameter omni-modal model, successor to MiniCPM-o 2.6, built from a SigLip2 vision encoder, a Whisper-medium audio encoder, a CosyVoice2 speech decoder, and a Qwen3-8B language backbone. It processes images, video, and audio and can generate spoken output, adding full-duplex live streaming so it can see, listen, and speak at once, plus a proactive mode that lets it comment unprompted, and bundled instruct and thinking modes. At this size it needs a high-end consumer GPU once quantized. Context length is 40,960 tokens. It is released under the Apache 2.0 license, permitting unrestricted commercial and research use, and was published in February 2026. Its defining upgrade over 2.6 is full-duplex interaction: earlier omni models process input and generate output in alternating turns, while 4.5 streams both directions at once without blocking.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M · 8K)Best on 24 GB (Q4_K_M · ~105 tok/s · 40K)Macs from 16 GB (Q4_K_M · ~23 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run MiniCPM O 4 5
One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.
Recommended
ollama run hf.co/openbmb/MiniCPM-o-4_5-gguf:Q4_K_MQ4_K_M fits fully on the RTX 4090 · 15.7 GB free · 16K context
Download: MiniCPM-o-4_5-Q4_K_M.gguf5.03 GBopenbmb
Ways to runOther ways
each opens step-by-stepRun in cloud
Your own private copy · Q4_K_M · 24 GB GPU
- The easiest way to try this model
- Running in a few minutes, easy setup
- Chat right here or connect any app
$5+ bonus for new users
Runs on RTX 3080 Ti (12 GB) (3 GB headroom) · Q4_K_M
- Generation speed
- ~95 tok/s
- generation speed
- Cost per 1M output tokens
- $0.36
- per 1M output tokens
How Much VRAM Does MiniCPM O 4 5 Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 4.6 GB | 9.1 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 40K context · ~143 tok/s | |
| 5.2 GB | 9.7 GB | 3-bit medium — 3-bit medium quantization | Fits: 40K context · ~127 tok/s | |
| 5.3 GB | 9.8 GB | 4-bit · legacy — 4-bit legacy quantization | Fits: 40K context · ~124 tok/s | |
| 5.9 GB | 10.4 GB | 4-bit small — 4-bit small quantization | Fits: 40K context · ~112 tok/s | |
| 6.2 GB | 10.8 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 40K context · ~105 tok/s | |
| 7.3 GB | 11.8 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 40K context · ~90 tok/s | |
| 8.3 GB | 12.9 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 40K context · ~79 tok/s | |
| 10.0 GB | 14.5 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 40K context · ~66 tok/s |
Selected Q4_K_M
4-bit medium quantization8.3GBat 16K context
- Weights
- 5.6 GB
- KV cache
- 2.4 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 15.7 GB free · ~105 tok/s
≥50 tok/s needs ≥479 GB/s · RTX 5090 ~187, RTX 3090 Ti ~105
Which Hardware Can Run MiniCPM O 4 5?
Q4_K_M · 6.2 GBMiniCPM O 4 5 (Q4_K_M) requires 6.2 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 9+ GB is recommended. Using the full 41K context window can add up to 5.7 GB, bringing total usage to 12.0 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 3070 Ti. 151 devices with unified memory can run MiniCPM O 4 5, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), NVIDIA Jetson Orin Nano 8GB (Super).
12 GB
12 GB tier →fits · 3.7 GB free · up to 40K ctx
10 GPUs · 1 device family
RTX 3080 Ti~95Simulate output on RTX 3080 TiRTX 5070~70Simulate output on RTX 5070RTX 4070~53Simulate output on RTX 4070iPhone~9Simulate output on Apple iPhone 17 Pro20 GB
fits · 11.7 GB free · up to 40K ctx
1 GPU
RX 7900 XT~77Simulate output on RX 7900 XT32–40 GB
fits · 23.7 GB free · up to 40K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~187Simulate output on RTX 5090M5 Max36 GB~52Simulate output on Mac Studio M5 Max (36 GB)M532 GB~17Simulate output on MacBook Air 13" M5 (32 GB)Orin~21Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~14Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Alternatives to MiniCPM O 4 5
MiniCPM family · 18 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| SenseNova Loop 9B Previewsensenova | — | 8.2 GB | ~103 | same size |
| Granite Switch 4.1 8BIBM | — | 8.4 GB | ~111 | same size |
| ZDTaichu5.0 9BTaichuAI | — | 9.2 GB | ~101 | same size |
| Ornith 1.5 9Bornith-ai | — | 6.6 GB | ~103 | same size |
| Apertus V1.5 8Bswiss-ai | — | 8.5 GB | ~111 | same size |
| ZAYA1 VL 8BZyphra | — | 7.5 GB | ~346 | same size |
Specifications
- Publisher
- OpenBMB
- Family
- MiniCPM
- Parameters
- 9.4B
- Architecture
- MiniCPMO
- Context Length
- 40,960 tokens
- Vocabulary Size
- 151,748
- Release Date
- 2026-02-03
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does MiniCPM O 4 5 need?
MiniCPM O 4 5 requires 6.2 GB of VRAM at Q4_K_M, or 19.4 GB at BF16. Full 41K context adds up to 5.7 GB (12.0 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 9.4B × 4.8 bits ÷ 8 = 5.6 GB
KV Cache + Overhead ≈ 0.6 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
KV Cache + Overhead ≈ 6.4 GB (at full 41K context)
VRAM usage by quantization
What's the best quantization for MiniCPM O 4 5?
For MiniCPM O 4 5, Q4_K_M (6.2 GB) offers the best balance of quality and VRAM usage. Q5_0 (6.5 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 4.6 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run MiniCPM O 4 5 on a Mac?
Yes — Mac Mini M2 Pro (16 GB) and 133 other Macs can run MiniCPM O 4 5. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 6.2 GB of usable unified memory (RAM minus macOS overhead).
Can I run MiniCPM O 4 5 locally?
Yes — MiniCPM O 4 5 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is MiniCPM O 4 5?
At Q4_K_M, MiniCPM O 4 5 can reach ~771 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~105 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 6.23 × 0.65 = ~835 tok/s
Estimated speed at Q4_K_M (6.2 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of MiniCPM O 4 5?
At Q4_K_M, the download is about 5.62 GB. The full-precision BF16 version is 18.74 GB. The smallest option (Q2_K) is 3.98 GB.
Which GPUs can run MiniCPM O 4 5?
52 consumer GPUs can run MiniCPM O 4 5 at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT, AMD Radeon RX 7600. 40 GPUs have plenty of headroom for comfortable inference.
Which devices can run MiniCPM O 4 5?
151 devices with unified memory can run MiniCPM O 4 5 at Q4_K_M (6.2 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.