SILMA 9B
A 9B open-weight Arabic specialist built on Google Gemma.
About
SILMA 9B (v1.0) is a 9-billion-parameter Arabic-specialized model from Silma AI, built on Google's Gemma and tuned for Arabic language nuance and culture. It topped the Open Arabic LLM Leaderboard v1 at release and is distributed as open weights with an instruct variant (silma-ai/SILMA-9B-Instruct-v1.0), spawning community fine-tunes such as Moroccan Darija variants.
Limitations: Gemma-based licensing rather than Apache 2.0 — review the Gemma terms; vendor benchmark claims are self-reported; no accompanying research paper was published.
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 — 6/10
Scored 6/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 67; Trending score 60/100 from researcher assessment.