Kanana-1.5-v-3B
Kakao's open-source 3B multimodal model for Korean and English image understanding.
About
Kanana-1.5-v-3B is a lightweight multimodal language model released open-source by Kakao's Unified Foundation Model task force on July 24, 2025. It processes both text and image inputs with 3.6B total parameters, combining an image encoder with the kanana-1.5-3b language model.
The model targets image captioning, document understanding, OCR-based reasoning and multimodal instruction following in English and Korean, with a 32k context window.
Released under the Kanana license, it followed Kakao's broader Kanana open-source releases and was positioned for bilingual Korean-English environments.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 116; Trending score 72/100 from researcher assessment.