MiniCPM5 2B Base — Hardware Requirements & GPU Compatibility
Chat
MiniCPM5 2B Base is a 2.5B-parameter open language model from OpenBMB in the MiniCPM family. It supports a context window of up to 524,288 tokens. At Q4_K_M it needs about 1.90 GB of VRAM — see which GPUs and Macs can run it below.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M · 128K)Best on 24 GB (Q4_K_M · ~345 tok/s · 256K)Macs from 8 GB (Q4_K_M · ~38 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run MiniCPM5 2B Base
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
vllm serve openbmb/MiniCPM5-2B-Base --max-model-len 16384vLLM repo with its own quantisation
Download: Weights on Hugging Face
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3080 Ti (12 GB) (9 GB headroom) · Q4_K_M
- Generation speed
- ~312 tok/s
- generation speed
- Cost per 1M output tokens
- $0.11
- per 1M output tokens
How Much VRAM Does MiniCPM5 2B Base Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 1.5 GB | 2.8 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 512K context · ~449 tok/s | |
| 1.6 GB | 2.9 GB | 3-bit medium — 3-bit medium quantization | Fits: 256K context · ~407 tok/s | |
| 1.9 GB | 3.2 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 256K context · ~345 tok/s | |
| 2.2 GB | 3.5 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 256K context · ~301 tok/s | |
| 2.5 GB | 3.8 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 256K context · ~266 tok/s | |
| 2.9 GB | 4.2 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 256K context · ~226 tok/s | |
| 5.4 GB | 6.7 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Fits: 256K context · ~121 tok/s |
Selected Q4_K_M
4-bit medium quantization2.5GBat 16K context
- Weights
- 1.5 GB
- KV cache
- 0.7 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 21.5 GB free · ~345 tok/s
≥50 tok/s needs ≥146 GB/s · RTX 5090 ~613, RTX 3090 Ti ~345
Which Hardware Can Run MiniCPM5 2B Base?
Q4_K_M · 1.9 GBMiniCPM5 2B Base (Q4_K_M) requires 1.9 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 3+ GB is recommended. Using the full 524K context window can add up to 22.5 GB, bringing total usage to 24.4 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run MiniCPM5 2B Base, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).
≤ 11 GB
≤ 11 GB tier →fits · 5.5 GB free · up to 128K ctx
16 GPUs · 1 Mac class · 2 device families
RTX 3080~260Simulate output on RTX 3080RTX 2080 Ti~211Simulate output on RTX 2080 TiRTX 3070 Ti~208Simulate output on RTX 3070 TiM3*8 GB~38Simulate output on iMac 24" M3 (8 GB)Orin Nano~35Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~25Simulate output on iPhone 1720 GB
fits · 17.5 GB free · up to 256K ctx
1 GPU
RX 7900 XT~253Simulate output on RX 7900 XT32–40 GB
fits · 29.5 GB free · up to 512K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~613Simulate output on RTX 5090M5 Max36 GB~170Simulate output on Mac Studio M5 Max (36 GB)M532 GB~57Simulate output on MacBook Air 13" M5 (32 GB)Orin~70Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~46Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Alternatives to MiniCPM5 2B Base
MiniCPM family · 18 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| WeVisOne VL 2BTencent | — | 3.6 GB | ~328 | same size |
| North Micro Vision InstructCohere | — | 3.7 GB | ~323 | same size |
| LFM2.5 2.6BLiquid AI | — | 2.3 GB | ~321 | same size |
| Granite 3.1 2B InstructIBM | — | 3.2 GB | ~329 | same size |
| Gemma 2 2B ITGoogle | — | 2.5 GB | ~379 | same size |
| Ouro 2.6B ThinkingByteDance | — | 12.1 GB | ~102 | same size |
Fine-tunes & variants(1)
Specifications
- Publisher
- OpenBMB
- Family
- MiniCPM
- Parameters
- 2.5B
- Architecture
- LlamaForCausalLM
- Context Length
- 524,288 tokens
- Vocabulary Size
- 130,560
- Release Date
- 2026-08-27
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does MiniCPM5 2B Base need?
MiniCPM5 2B Base requires 1.9 GB of VRAM at Q4_K_M, or 5.4 GB at BF16. Full 524K context adds up to 22.5 GB (24.4 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 2.5B × 4.8 bits ÷ 8 = 1.5 GB
KV Cache + Overhead ≈ 0.4 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 ≈ 22.9 GB (at full 524K context)
VRAM usage by quantization
What's the best quantization for MiniCPM5 2B Base?
For MiniCPM5 2B Base, Q4_K_M (1.9 GB) offers the best balance of quality and VRAM usage. Q5_K_M (2.2 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 1.5 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run MiniCPM5 2B Base on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run MiniCPM5 2B Base. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 1.9 GB of usable unified memory (RAM minus macOS overhead).
Can I run MiniCPM5 2B Base locally?
Yes — MiniCPM5 2B Base can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is MiniCPM5 2B Base?
At Q4_K_M, MiniCPM5 2B Base can reach ~2526 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~345 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 ÷ 1.9 × 0.65 = ~2737 tok/s
Estimated speed at Q4_K_M (1.9 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of MiniCPM5 2B Base?
At Q4_K_M, the download is about 1.51 GB. The full-precision BF16 version is 5.03 GB. The smallest option (Q2_K) is 1.07 GB.
Which GPUs can run MiniCPM5 2B Base?
52 consumer GPUs can run MiniCPM5 2B Base at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 52 GPUs have plenty of headroom for comfortable inference.
Which devices can run MiniCPM5 2B Base?
163 devices with unified memory can run MiniCPM5 2B Base at Q4_K_M (1.9 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.