---
title: MiniCPM5-2B-OptiQ-4bit
canonical_url: "https://www.modelscope.cn/models/latentnode/MiniCPM5-2B-OptiQ-4bit"
md_url: "https://www.modelscope.cn/models/latentnode/MiniCPM5-2B-OptiQ-4bit.md"
repository: latentnode/MiniCPM5-2B-OptiQ-4bit
chinese_name: MiniCPM5-2B-OptiQ-4bit
last_updated: 2026-09-11
license: apache-2.0
pipeline_tag: text-generation
tasks:
- text-generation
model_type:
- llama
architectures:
- LlamaForCausalLM
base_model:
- openbmb/MiniCPM5-2B
base_model_relation: quantized
parameters: 520.9M
tensor_type:
- BF16
- U32
library_name:
- mlx
- safetensors
language:
- en
- zh
downloads: 5
stars: 0
tags:
- mlx
- mlx-optiq
- optiq
- minicpm
- minicpm5
- llama
- text-generation
- quantization
- mixed-precision
- apple-silicon
- long-context
- tool-calling
- on-device
- edge-ai
---
# MiniCPM5-2B-OptiQ-4bit
> MiniCPM5-2B-OptiQ-4bit - latentnode 在 ModelScope 开源的模型。MiniCPM5-2B-OptiQ-4bit
latentnode/MiniCPM5-2B-OptiQ-4bit 是 ModelScope 魔搭社区上的 520.9M 参数text-generation模型，采用 apache-2.0 许可，基于 openbmb/MiniCPM5-2B 构建。
- **Repository**: latentnode/MiniCPM5-2B-OptiQ-4bit
- **License**: apache-2.0
- **Tasks**: text-generation
- **Parameters**: 520.9M
- **Base model**: openbmb/MiniCPM5-2B
- **Tags**: mlx, mlx-optiq, optiq, minicpm, minicpm5, llama, text-generation, quantization, mixed-precision, apple-silicon, long-context, tool-calling, on-device, edge-ai
- **Downloads**: 5
- **Stars**: 0
- **Last updated**: 2026-09-11
Source: https://www.modelscope.cn/models/latentnode/MiniCPM5-2B-OptiQ-4bit
---
# MiniCPM5-2B-OptiQ-4bit
> **Built with [mlx-optiq](https://mlx-optiq.com)**, the MLX-native toolkit to quantize, prune, fine-tune, and serve LLMs locally on Apple Silicon, no PyTorch and no cloud. [Try the Lab](https://mlx-optiq.com/docs/lab/) · [All OptiQ quants](https://mlx-optiq.com/models) · [Docs](https://mlx-optiq.com/docs/)
Mixed-precision MLX quantization of [openbmb/MiniCPM5-2B](https://huggingface.co/openbmb/MiniCPM5-2B), OpenBMB's 2B-class Llama-architecture model with hybrid reasoning, tool calling and a 128k context. Apache-2.0 from the base model.
- **Base:** openbmb/MiniCPM5-2B (42 layers, Llama architecture, hybrid reasoning, 128k context)
- **Quantization:** OptiQ sensitivity-aware mixed precision (4-bit + 8-bit)
- **Disk size:** 1,837 MB
- **Calibration:** the bundled `optiq.jsonl` mix (40 samples across prose, reasoning, code, agent loops, tool-calling, instructions)
The OptiQ optimizer measured every linear layer against the bf16 model and kept **112 of 295 at 8-bit** (the ones whose output moved most under 4-bit: output projections, the last blocks, lm_head) and **183 at 4-bit**. The full per-layer assignment is in `optiq/metadata.json` next to the safetensors.
## Capability Score
Six benchmarks, one number, the same recipe as every mlx-optiq quant (non-thinking mode, for cross-family comparability).
| Benchmark | Score |
|---|---:|
| MMLU (5-shot, 1000) | 59.8% |
| GSM8K (1000, 3-shot CoT) | 82.1% |
| IFEval (full set, strict) | 86.7% |
| BFCL V3 (simple AST, 200 calls) | 82.5% |
| HumanEval (164, pass@1) | 81.1% |
| HashHop (long-context retrieval) | 24.0% |
| **Capability Score** | **69.36** |
For scale: this 1.8 GB file scores level with the 2.8 GB [Qwen3.5-4B-OptiQ-4bit](https://huggingface.co/mlx-community/Qwen3.5-4B-OptiQ-4bit) (68.76), ahead on instruction following, GSM8K and HumanEval, behind on MMLU and tool-call format.
## Loading
```python
from mlx_lm import load, generate
model, tok = load("mlx-community/MiniCPM5-2B-OptiQ-4bit")
prompt = tok.apply_chat_template(
[{"role": "user", "content": "Summarize the plot of The Iliad in three sentences."}],
tokenize=False,
add_generation_prompt=True,
enable_thinking=False,
)
print(generate(model, tok, prompt=prompt, max_tokens=300))
```
## Hybrid reasoning: think or no-think
The chat template takes an `enable_thinking` flag. On, the model emits a `...` block before answering, which helps on math, planning and multi-step code. The scores above are with thinking **off**.
| Mode | temperature | top_p | Use when |
|---|---:|---:|---|
| No-think (default) | 0.7 | 0.95 | Fast assistant, rewriting, conversational |
| Think | 0.9 | 0.95 | Math, code, multi-hop reasoning |
`optiq serve` forwards `chat_template_kwargs` verbatim, and also accepts the model name with a `:think` / `:no-think` suffix:
```bash
optiq serve --model mlx-community/MiniCPM5-2B-OptiQ-4bit --port 8000
curl -s http://localhost:8000/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{"model":"mlx-community/MiniCPM5-2B-OptiQ-4bit:think",
"messages":[{"role":"user","content":"What is 17 * 23?"}]}'
```
## Tool calling
MiniCPM5 uses its own `` call format. `optiq serve` parses it into OpenAI `tool_calls`, so agents that speak the OpenAI or Anthropic protocol (OptiQ Code, Claude Code via the Anthropic endpoint, OpenCode, Codex) work unchanged.
## KV cache
`kv_config.json` in this repo is a per-layer mixed-precision KV-cache plan from `optiq kv-cache`. Pass it to `optiq serve --kv-config` for a smaller cache at long context without the quality loss of uniform 4-bit KV.
## Fine-tuning
Small enough to LoRA-train on a 24 GB Mac at `max_seq_length=2048`. The [sensitivity-aware LoRA](https://mlx-optiq.com/blog/sensitivity-aware-lora) overlay reads `optiq/metadata.json` and gives 8-bit layers twice the adapter rank of 4-bit layers at the same parameter budget.
```bash
optiq lora train mlx-community/MiniCPM5-2B-OptiQ-4bit \
--data ./my_training_data \
--preset default \
--max-seq-length 2048
```
## Reproducibility
```bash
optiq convert openbmb/MiniCPM5-2B --target-bpw 5.0 --candidate-bits 4,8 --reference bf16
```
Deterministic given the same calibration mix.
## Quantize your own
```bash
pip install mlx-optiq
optiq convert --target-bpw 5.0 --candidate-bits 4,8
optiq lab # local workbench: chat, compare, quantize, fine-tune
```
## License + provenance
Apache-2.0, inherited from [openbmb/MiniCPM5-2B](https://huggingface.co/openbmb/MiniCPM5-2B). Quantized with [mlx-optiq](https://mlx-optiq.com).