DataLearner logo
MI

MiniCPM5-2B

Reasoning modelTool useMiniCPM

MiniCPM5-2B

Also known as: MiniCPM 5 2B / MiniCPM5 2B

Release date: 2026-09-07Views: 3
Parameters
2.517B
Context length
128K
Multilingual
中文、英文 languages
Reasoning ability
No data

MiniCPM5-2B is an AI model published by OpenBMB, released on 2026-09-07, for Reasoning model, with 2.517B parameters, and 128K context length, with a 97.10 score on τ²-Bench - Telecom.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

MiniCPM5-2B

Model basics

Reasoning traces
Supported
Thinking modes
Thinking Mode (Default)Standard Mode
Context length
128K tokens
Max output length
No data
Model type
Reasoning model
Modality (in / out)
Text → Text
Supported languages
Chinese, English (中文、英文 languages)
Release date
2026-09-07
Model file size
No data
MoE architecture
No
Total params / Active params
2.517B / Not applicable
Knowledge cutoff
No data
MiniCPM5-2B

Open source & experience

Code license
Weights license
Apache 2.0- Commercial use permitted
MiniCPM5-2B

Official resources

DataLearnerAI blog
N/A
MiniCPM5-2B

API details

API speed
No data
No public API pricing yet.
MiniCPM5-2B

Benchmark Results

MiniCPM5-2B currently shows benchmark results led by τ²-Bench - Telecom (11 / 36, score 97.10), AIME2025 (47 / 107, score 86.50), LiveCodeBench (59 / 128, score 69.10). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking

General Knowledge

2 evaluations
Benchmark / mode
Score
Rank/total
MMLU Pro
unknown
70.80
94 / 134
HLE
unknown
8.90
168 / 190

General Evaluation

1 evaluations
Benchmark / mode
Score
Rank/total
70.20
195 / 271

Coding and Software Engineer

4 evaluations
Benchmark / mode
Score
Rank/total
69.10
59 / 128
46.40
103 / 116
SciCode
unknown
26.30
16 / 16
14.40
61 / 61

Math and Reasoning

4 evaluations
Benchmark / mode
Score
Rank/total
MATH-500
unknown
94.60
26 / 45
AIME2025
unknown
86.50
47 / 107
AIME 2026
unknown
86.50
19 / 21
63.80
2 / 2

Agent Level Benchmark

2 evaluations
Benchmark / mode
Score
Rank/total
97.10
11 / 36
20.80
10 / 13

Instruction Following

2 evaluations
Benchmark / mode
Score
Rank/total
IFEval
unknown
86.70
2 / 2
IF Bench
unknown
66.30
24 / 36

Long Context

2 evaluations
Benchmark / mode
Score
Rank/total
AA-LCR
unknown
59
27 / 29
43.70
14 / 14

AI Agent - Tool Usage

2 evaluations
Benchmark / mode
Score
Rank/total
BFCL-V4
unknown
66.60
1 / 2
8.60
50 / 50

Productivity Knowledge

1 evaluations
Benchmark / mode
Score
Rank/total
19.60
26 / 26
MiniCPM5-2B

Publisher

MiniCPM5-2B

Model Overview

MiniCPM5-2B is an AI model published by OpenBMB, released on 2026-09-07, for Reasoning model, with 2.517B parameters, and 128K context length, with a 97.10 score on τ²-Bench - Telecom.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code