Skip to content
← Models
ClosedQCRI / Hamad Bin Khalifa University/10 trustMetrics-checked

Fanar 27B

Qatar's sovereign Arabic AI — 27B Arabic-centric model, fully built in-house.

Visit official page →

About

Fanar is Qatar's sovereign Arabic-centric generative AI platform, built entirely at Hamad Bin Khalifa University's Qatar Computing Research Institute (QCRI) with no dependency on external AI providers. Fanar 2.0 (paper: March 2026, arXiv 2603.16397) centers on Fanar-27B — continual pretraining of Gemma-3-27B on ~120B curated tokens — with 32K context and native selective reasoning, delivering +9.1 pts Arabic knowledge and +7.3 pts ArabicMMLU over v1 despite 8x fewer tokens.

The full stack includes FanarGuard (4B bilingual safety filter), Aura speech models, Oryx Arabic-aware vision, Islamic RAG (Fanar-Sadiq), classical Arabic poetry (Fanar-Diwan), and agentic tool-calling — a resource-constrained sovereignty showcase (256 H100s) for a language with ~0.5% of web data.

Honest limits: weights are not broadly open — access is via Qatar's Fanar platform/apps rather than public checkpoints; Arabic-first means English/general capability trails frontier models; the ecosystem is nascent outside the Gulf.

Specs

Context window

Pricing

Free quota

API access

Weights

Closed — API only

Released

March 1, 2026

Benchmarks

Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.

Use cases

arabic-nlpchatmultilingualtranslationspeech

Trust score — 6/10

/10 trustMetrics-checked

Scored 6/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 55; Trending score 50/100 from researcher assessment.

More in the fanar family

Sources