OpenBMBReleased Sep 7, 2026
Open weights

MiniCPM5 2B

Its best results are in Knowledge Work (#118 of 186) and Reliability (#132 of 188); weakest in Programmatic Reasoning (#142 of 187). Three of seven dimensions are estimated, because no results are published yet.

Composite IQ ?
97
#152 of 186 ranked models · ahead of 18%
China (Other)'s #16 model, behind Ling 3.1 Flash (120)
Effective cost
Not priced
No standard API rate
Speed
Not measured
No speed data yet
Context
131Ktokens
131,072 tokens in
Evidence
4of 7 dims
11 of 39 benchmarks

IQ profile

Seven dimensions, averaged equally into Composite IQ
Compare shape with
100 avg human130160● Abstract87#137 · est.● Math88#170 · est.● Academic102#168● Coding99#142● Computer Use94#145 · est.● Reliability107#132● Knowledge Work102#118
MeasuredEstimated95% range
#1 in dimensionTop-10 median

IQ dimensions

Ranks are among models scored in each dimension. Ranges are model-based estimates, not calibrated confidence intervals.

DimensionIQPosition · 70–160RankEvidencevs Top 10
87estimated
#137 / 221
0/5
-45.5
88estimated
#170 / 186
0/5
-61.5
102measured
#168 / 223
4/6
-33
99measured
#142 / 187
1/6
-52
94estimated
#145 / 186
0/5
-52.5
107measured
#132 / 188
2/6
-28.5
102measured
#118 / 186
4/6
-30
MiniCPM5 2B95% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human

Benchmark results

11 of 39 Composite IQ benchmarks have a published result for MiniCPM5 2B. Benchmarks without a result stay listed so gaps are visible.

Abstract Reasoning IQ 87 · estimatedNo published results0/5
No published results yet for Abstract Reasoning. The 87 shown is estimated from MiniCPM5 2B's other dimensions and will be replaced as results come in. How estimates work
ARC-AGI-3ARC-AGI-2ARC-AGI-1Chess PuzzlesMystery Game Puzzles
Mathematical Reasoning IQ 88 · estimatedNo published results0/5
No published results yet for Mathematical Reasoning. The 88 shown is estimated from MiniCPM5 2B's other dimensions and will be replaced as results come in. How estimates work
FrontierMath Tier 4FrontierMath Tier 1-3ProofBench 1.1MathArenaAIME
Academic Reasoning IQ 102 · measuredBest: #132 of 159 on CritPt4/6
BenchmarkScorePercentileRankLeader · source
CritPt
Artificial Analysis v4.3.2 · Oct 27
0.3%
#132 / 159
Leader Muse Wonder Full · 32.9%
Artificial Analysis v4.3.2 · Oct 27
GPQA Diamond
Artificial Analysis v5.0.1 · Nov 13
70.2%
#127 / 147
Leader gpt-6-astra · 96.1%
Artificial Analysis v5.0.1 · Nov 13
Humanity's Last Exam
Artificial Analysis v4.3.2 · Oct 27
8.9%
#149 / 167
Leader opus-5.5 · 61.4%
Artificial Analysis v4.3.2 · Oct 27
SciCode
Artificial Analysis v4.3.2 · Oct 27
26.3%
#159 / 167
Leader opus-5.5 · 66.9%
Artificial Analysis v4.3.2 · Oct 27
Not yet reported
MMLU-ProMMMU-Pro
Programmatic Reasoning IQ 99 · measuredBest: #97 of 109 on Terminal-Bench 4.01/6
BenchmarkScorePercentileRankLeader · source
Terminal-Bench 4.0
Artificial Analysis v5.0.1 · Nov 13
0.0%
#97 / 109
Leader Kimi K3.1 Pro · 67.2%
Artificial Analysis v5.0.1 · Nov 13
Not yet reported
LiveCodeBenchIOI V2ProgramBenchSWE-rebenchMirrorCode
Computer Use IQ 94 · estimatedNo published results0/5
No published results yet for Computer Use. The 94 shown is estimated from MiniCPM5 2B's other dimensions and will be replaced as results come in. How estimates work
BrowseCompOSWorld-VerifiedMCP AtlasArena.ai Agent ArenaAgents' Last Exam
Reliability IQ 107 · measuredBest: #97 of 158 on AA Omniscience2/6
BenchmarkScorePercentileRankLeader · source
AA Omniscience
Artificial Analysis v4.3.2 · Oct 27
-11.6
#97 / 158
Leader opus-5.5 · 46.4
Artificial Analysis v4.3.2 · Oct 27
AA Long Context Reasoning v1.1
Artificial Analysis v4.3.2 · Oct 27
59.3%
#130 / 154
Leader kimi-k3 · 88.7%
Artificial Analysis v4.3.2 · Oct 27
Not yet reported
SimpleQA VerifiedBullshitBench v2IFBenchFACTS Grounding
Knowledge Work IQ 102 · measuredBest: #96 of 124 on GDPval-AA v2.1 normalized score4/6
BenchmarkScorePercentileRankLeader · source
GDPval-AA v2.1 normalized score
Artificial Analysis v4.3.2 · Oct 23
10.4%
#96 / 124
Leader opus-5.5 · 68.3%
Artificial Analysis v4.3.2 · Oct 23
AA-Briefcase v1.1 Elo
Artificial Analysis v5.0.1 · Nov 13
423
#96 / 116
Leader sonnet-5.5 · 1,816
Artificial Analysis v5.0.1 · Nov 13
AutomationBench-AA
Artificial Analysis v5.0.1 · Nov 13
1.8%
#96 / 112
Leader Gemini 4 Carbon · 77.8%
Artificial Analysis v5.0.1 · Nov 13
GDP.pdf
Artificial Analysis v5.0.1 · Nov 13
1.4%
#103 / 111
Leader Muse Wonder Full · 32.6%
Artificial Analysis v5.0.1 · Nov 13
Not yet reported
AA-AnalystAgentEnterpriseOps-Gym-AA

Release & availability

Facts from the publisher's announcement and documentation.

Released
Developer
OpenBMB
Access
Apache 2.0 open weights on Hugging Face.
Weights
Open weightsDownloadable and self-hostable
Inputs
text
Context
131,072 tokens
API price
No per-token rateNo current standard hosted token price has been verified for this checkpoint; downloadable weights are not a zero-dollar hosted API.

What OpenBMB says

  • Compact dense model for local assistants, coding, tool use, and reasoning.
  • Final checkpoint includes reinforcement learning and on-policy distillation; base, SFT, and quantized artifacts are separate training/runtime variants.

Limitations

  • No verified nonzero hosted API pricing.

Publisher statements, not AI IQ measurements.

About this score

How much of MiniCPM5 2B's IQ is measured, and how the rest is estimated.

11/ 39benchmark results
4/ 7dimensions measured
3estimated
0/5Abstract
0/5Math
4/6Academic
1/6Coding
0/5Computer Use
2/6Reliability
4/6Knowledge Work
Bar height = share of benchmarks with resultsEstimatedNo score
Measured results are shown as published

Every score in the benchmark tables comes from its source unchanged. Composite IQ averages all seven dimensions equally.

Estimates fill the gaps

Abstract Reasoning, Mathematical Reasoning and Computer Use have no published results yet, so their IQ is estimated from MiniCPM5 2B's other dimensions.

A missing result is not a zero

Benchmark versions and evaluation harnesses can differ between sources, so small gaps between models may not be meaningful.

Read the scoring methodology →

Compare MiniCPM5 2B

Side-by-side pages for the models it is most often weighed against.

MiniCPM5 2BComparison model
Browse all comparisons →

Frequently asked questions

Quick answers about MiniCPM5 2B, drawn from the same data as the rest of this page.

How AI IQ scores models →Get this data from the API →
What is the IQ of MiniCPM5 2B?
MiniCPM5 2B has an estimated IQ of 97, which ranks #152 out of 186 public models on the AI IQ leaderboard. The score combines seven equally weighted reasoning dimensions using measured benchmark results and conservative estimates for missing coverage.
What is MiniCPM5 2B best at?
Among the dimensions with published benchmark results, MiniCPM5 2B ranks highest in Knowledge Work (IQ 102, #118 of 186) and Reliability (IQ 107, #132 of 188). Its lowest-ranked measured dimension is Programmatic Reasoning (IQ 99, #142 of 187).
What is the EQ of MiniCPM5 2B?
MiniCPM5 2B doesn't have an EQ score yet. EQ (emotional reasoning) is calculated from EQ-Bench 4 and AttuneBench, and neither has published a result for MiniCPM5 2B. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.
Which benchmarks has MiniCPM5 2B been tested on?
MiniCPM5 2B has published results on 11 of the 39 benchmarks behind AI IQ's Composite IQ, including AA Omniscience, GDPval-AA v2.1 normalized score, AA-Briefcase v1.1 Elo, CritPt and AA Long Context Reasoning v1.1.
How much does MiniCPM5 2B cost?
AI IQ doesn't have a verified standard per-token API price for MiniCPM5 2B, so it can't calculate an effective cost. This is common for models offered only through subscriptions, restricted-access programs, or self-hosting.
Is MiniCPM5 2B open source?
Yes — MiniCPM5 2B is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.
Who developed MiniCPM5 2B?
MiniCPM5 2B was developed by OpenBMB, based in China.
When was MiniCPM5 2B released?
MiniCPM5 2B was publicly released on September 7, 2026.
How large is the context window of MiniCPM5 2B?
MiniCPM5 2B supports a context window of 131,072 tokens — the amount of text it can consider in a single request.
How fast is MiniCPM5 2B?
AI IQ doesn't have independent speed measurements for MiniCPM5 2B yet. Speed comes from independent tests of a model's hosted API, which aren't available for every model — especially new, restricted-access, or self-hosted ones.