FA
Fanar-2-27B
QCRI / HBKU
All models
About
Arabic-centric flagship of the Fanar 2.0 release (Mar 2026), continually pretrained from google/gemma-3-27b-pt on ~166B Arabic, English and code tokens with 32K context. It adds native Arabic reasoning traces, selective thinking mode and tool calling. This repo is text-in/text-out; image generation, image understanding and poetry are separate Fanar-2 models.
Community evaluations
No evaluations yet
Be the first to say whether it actually works: real Saudi audio, real Gulf Arabic text, not just the benchmark numbers.
Evaluate this model
What did you actually test? Real audio, real text, which language? Evaluations are published after moderation.