Is bartowski/MiniCPM5-2B-GGUF good for agentic (tool-using) tasks?

In our benchmarks, bartowski/MiniCPM5-2B-GGUF scores 52.8/100 for agentic workflows. It runs at about 166.8 tok/s, so if you want more speed, unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q4_K_M is faster (~303.4 tok/s) and still scores well for agentic workflows (82.6/100).

How fast is bartowski/MiniCPM5-2B-GGUF for local inference?

Across 10 community benchmark runs, bartowski/MiniCPM5-2B-GGUF reaches up to 235.0 tok/s and averages 166.8 tok/s, with the fastest results on NVIDIA GeForce RTX 4070.

How much memory does bartowski/MiniCPM5-2B-GGUF need?

The leanest observed configuration used about 5.2 GB of memory (quantizations tested: Q4_K_M, Q6_K_L).

Which tools have been used to run bartowski/MiniCPM5-2B-GGUF?

Benchmarks were submitted using Unsloth Studio, llama.cpp. Results are community-contributed and updated as new runs arrive.

Model benchmarks

bartowski/MiniCPM5-2B-GGUF local LLM performance

As of October 2026, bartowski/MiniCPM5-2B-GGUF runs at up to 235.0 tok/s for local inference (best of 10 community benchmark runs across 2 GPUs).

Unsloth Studiollama.cppQ4_K_MQ6_K_L

ShareReddit X

Model size

Peak speed

235.0 tok/s

Average speed

166.8 tok/s

Avg PP

2505.3 tok/s

Min memory

5.2 GB

Max context

131,072 tokens

Avg output / run

33,535 tokens

Avg runtime / run

4m 39s

Avg quality

46.1

Benchmark runs

GPUs tested

Quality by task

Average LLM-judged quality (0–100) with the run-to-run spread shown as a P5–P95 band, overall and for each benchmark task, across all 10 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.

Task	P5 (low)	Avg	P95 (high)
Overall	35.5	46.1	52.5
Agent Workflow	30.9	52.8	72.2
Code Generation	4.0	16.0	29.6
Role Play & Narrative	20.4	44.9	61.6
Research & Analysis	55.6	70.8	78.4

Performance by hardware and tool

Every hardware/tool/quantization combination bartowski/MiniCPM5-2B-GGUF has been benchmarked on, ranked by peak token generation speed. Last updated October 2026.

Hardware	Tool	Quant	Best tok/s	Avg tok/s	Memory	Context	Quality	Runs
NVIDIA GeForce RTX 4070	Unsloth Studio	Q4_K_M	235.0 tok/s	183.0 tok/s	5.2 GB	131,072 tokens	43.6	4
NVIDIA GeForce RTX 4070	llama.cpp	—	224.7 tok/s	205.8 tok/s	5.3 GB	131,072 tokens	48.3	3
NVIDIA GeForce RTX 4070	Unsloth Studio	Q6_K_L	166.1 tok/s	147.6 tok/s	7.3 GB	131,072 tokens	43.9	2
Intel(R) Arc(TM) 140V GPU (16GB)	Unsloth Studio	—	23.2 tok/s	23.2 tok/s	n/a	65,536 tokens	53.7	1

Benchmark runs

All 10 bartowski/MiniCPM5-2B-GGUF runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.

Open run page →

Frequently asked questions

Is bartowski/MiniCPM5-2B-GGUF good for agentic (tool-using) tasks?: In our benchmarks, bartowski/MiniCPM5-2B-GGUF scores 52.8/100 for agentic workflows. It runs at about 166.8 tok/s, so if you want more speed, unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q4_K_M is faster (~303.4 tok/s) and still scores well for agentic workflows (82.6/100).
How fast is bartowski/MiniCPM5-2B-GGUF for local inference?: Across 10 community benchmark runs, bartowski/MiniCPM5-2B-GGUF reaches up to 235.0 tok/s and averages 166.8 tok/s, with the fastest results on NVIDIA GeForce RTX 4070.
How much memory does bartowski/MiniCPM5-2B-GGUF need?: The leanest observed configuration used about 5.2 GB of memory (quantizations tested: Q4_K_M, Q6_K_L).
Which tools have been used to run bartowski/MiniCPM5-2B-GGUF?: Benchmarks were submitted using Unsloth Studio, llama.cpp. Results are community-contributed and updated as new runs arrive.