Multimodal Inference Optimization SGLang

MiniCPM-V 4.6

OpenBMB multimodal model release

Visit site

OpenBMB's MiniCPM-V 4.6 multimodal model release.

Pricing

Official site · May 12, 2026, 6:30 AM

Pricing notes were collected, but there are no normalized numeric fields to display yet.

Presented as open-source/self-hosted in official model materials; no public price list or rate card was identified.

OpenBMB's official materials indicate MiniCPM-V 4.6 is an open multimodal model, but I could not confirm any public paid API or token-based pricing from first-party sources. No normalized public price was found.

View source

Model Intelligence

Arena ranking

Benchmarkable

Yes

Model level

release

Intelligence Index

4.2

Coding Index

0.7

GPQA

0.31

HLE

0.05

SciCode

0.02

IFBench

0.27

LCR

0.06

TerminalBench Hard

TAU2

0.88

Recent stories

1 linked story

releasePRIMARY2026-05-11

OpenBMB releases MiniCPM-V 4.6 1.3B with 75.7 ms TTFT and 19x token efficiency

OpenBMB released MiniCPM-V 4.6 1.3B, claiming 55.8 percent lower vision-encoding FLOPs, 75.7 ms TTFT on a 4090, and about 1.5x token throughput over Qwen3.5 0.8B. It targets edge deployment across mobile platforms and common inference stacks.