MiniCPM-V 2.6
8B vision-language model that runs on your phone
About
MiniCPM-V 2.6 is OpenBMB's 8-billion-parameter vision-language model, open-sourced on August 6, 2024 under Apache 2.0. Built on SigLIP-400M and Qwen2-7B, it handles single images, multi-image reasoning and video, with strong OCR across 30+ languages. The team claims it outperforms GPT-4V on image and video understanding, which should be read as team-reported. It runs on iPad-class hardware and has official llama.cpp support, making it one of the more deployable open VLMs.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Team-reported: outperforms GPT-4V on single-image, multi-image and video understanding (OpenCompass average 65.2); state-of-the-art on OCRBench.
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Established OpenBMB release with official llama.cpp support; Apache 2.0 per project README; GPT-4V comparison claims are team-reported. License: Apache 2.0 (per project README)..
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 8.0/10
- Maintenance · recency, releases, issue load
- 8.0/10
- Openness · license + docs
- 8.0/10
- Manual review · sanity read + red-flag check
- 8.0/10
- · Established OpenBMB release with official llama.cpp support
- · Apache 2.0 per project README
- · GPT-4V comparison claims are team-reported