MiniCPM5-1B

A SOTA 1B on-device LLM, small yet powerful.

A dense 1B-parameter causal language-model checkpoint for on-device and resource-constrained local deployment, with hybrid reasoning and tool-use support.

Pricing

Model profile · Current snapshot

Output TPS

TTFT (s)

Model Intelligence

Context window

131,072 tokens

Arena ranking

Benchmarkable

Yes

Model level

release

Intelligence Index

11.9

GPQA

0.28

HLE

0.07

SciCode

0.04

IFBench

0.49

LCR

0.05

TerminalBench Hard

TAU2

0.81

Recent stories

1 linked story

releasePRIMARY2026-05-25

MiniCPM5-1B launches with 17.9 AA and ~0.5GB INT4 weights

OpenBMB released MiniCPM5-1B and says the model leads Artificial Analysis' small-model index at 17.9 while fitting into roughly 0.5GB in INT4. The release matters because it targets phones, browsers, and local runtimes with a sub-2B open model.