_version_ 1866912202257072128
author Fanar Team
Abbas, Ummar
Ahmad, Mohammad Shahmeer
Alam, Firoj
Altinisik, Enes
Asgari, Ehsannedin
Boshmaf, Yazan
Boughorbel, Sabri
Chawla, Sanjay
Chowdhury, Shammur
Dalvi, Fahim
Darwish, Kareem
Durrani, Nadir
Elfeky, Mohamed
Elmagarmid, Ahmed
Eltabakh, Mohamed
Fatehkia, Masoomali
Fragkopoulos, Anastasios
Hasanain, Maram
Hawasly, Majd
Husaini, Mus'ab
Jung, Soon-Gyo
Lucas, Ji Kim
Magdy, Walid
Messaoud, Safa
Mohamed, Abubakr
Mohiuddin, Tasnim
Mousi, Basel
Mubarak, Hamdy
Musleh, Ahmad
Naeem, Zan
Ouzzani, Mourad
Popovic, Dorde
Sadeghi, Amin
Sencar, Husrev Taha
Shinoy, Mohammed
Sinan, Omar
Zhang, Yifan
Ali, Ahmed
Kheir, Yassine El
Ma, Xiaosong
Ruan, Chaoyi
author_facet Fanar Team
Abbas, Ummar
Ahmad, Mohammad Shahmeer
Alam, Firoj
Altinisik, Enes
Asgari, Ehsannedin
Boshmaf, Yazan
Boughorbel, Sabri
Chawla, Sanjay
Chowdhury, Shammur
Dalvi, Fahim
Darwish, Kareem
Durrani, Nadir
Elfeky, Mohamed
Elmagarmid, Ahmed
Eltabakh, Mohamed
Fatehkia, Masoomali
Fragkopoulos, Anastasios
Hasanain, Maram
Hawasly, Majd
Husaini, Mus'ab
Jung, Soon-Gyo
Lucas, Ji Kim
Magdy, Walid
Messaoud, Safa
Mohamed, Abubakr
Mohiuddin, Tasnim
Mousi, Basel
Mubarak, Hamdy
Musleh, Ahmad
Naeem, Zan
Ouzzani, Mourad
Popovic, Dorde
Sadeghi, Amin
Sencar, Husrev Taha
Shinoy, Mohammed
Sinan, Omar
Zhang, Yifan
Ali, Ahmed
Kheir, Yassine El
Ma, Xiaosong
Ruan, Chaoyi
contents We present Fanar, a platform for Arabic-centric multimodal generative AI systems, that supports language, speech and image generation tasks. At the heart of Fanar are Fanar Star and Fanar Prime, two highly capable Arabic Large Language Models (LLMs) that are best in the class on well established benchmarks for similar sized models. Fanar Star is a 7B (billion) parameter model that was trained from scratch on nearly 1 trillion clean and deduplicated Arabic, English and Code tokens. Fanar Prime is a 9B parameter model continually trained on the Gemma-2 9B base model on the same 1 trillion token set. Both models are concurrently deployed and designed to address different types of prompts transparently routed through a custom-built orchestrator. The Fanar platform provides many other capabilities including a customized Islamic Retrieval Augmented Generation (RAG) system for handling religious prompts, a Recency RAG for summarizing information about current or recent events that have occurred after the pre-training data cut-off date. The platform provides additional cognitive capabilities including in-house bilingual speech recognition that supports multiple Arabic dialects, voice and image generation that is fine-tuned to better reflect regional characteristics. Finally, Fanar provides an attribution service that can be used to verify the authenticity of fact based generated content. The design, development, and implementation of Fanar was entirely undertaken at Hamad Bin Khalifa University's Qatar Computing Research Institute (QCRI) and was sponsored by Qatar's Ministry of Communications and Information Technology to enable sovereign AI technology development.
format Preprint
id arxiv_https___arxiv_org_abs_2501_13944
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Fanar: An Arabic-Centric Multimodal Generative AI Platform
Fanar Team
Abbas, Ummar
Ahmad, Mohammad Shahmeer
Alam, Firoj
Altinisik, Enes
Asgari, Ehsannedin
Boshmaf, Yazan
Boughorbel, Sabri
Chawla, Sanjay
Chowdhury, Shammur
Dalvi, Fahim
Darwish, Kareem
Durrani, Nadir
Elfeky, Mohamed
Elmagarmid, Ahmed
Eltabakh, Mohamed
Fatehkia, Masoomali
Fragkopoulos, Anastasios
Hasanain, Maram
Hawasly, Majd
Husaini, Mus'ab
Jung, Soon-Gyo
Lucas, Ji Kim
Magdy, Walid
Messaoud, Safa
Mohamed, Abubakr
Mohiuddin, Tasnim
Mousi, Basel
Mubarak, Hamdy
Musleh, Ahmad
Naeem, Zan
Ouzzani, Mourad
Popovic, Dorde
Sadeghi, Amin
Sencar, Husrev Taha
Shinoy, Mohammed
Sinan, Omar
Zhang, Yifan
Ali, Ahmed
Kheir, Yassine El
Ma, Xiaosong
Ruan, Chaoyi
Computation and Language
Artificial Intelligence
I.2.0; D.2.0
We present Fanar, a platform for Arabic-centric multimodal generative AI systems, that supports language, speech and image generation tasks. At the heart of Fanar are Fanar Star and Fanar Prime, two highly capable Arabic Large Language Models (LLMs) that are best in the class on well established benchmarks for similar sized models. Fanar Star is a 7B (billion) parameter model that was trained from scratch on nearly 1 trillion clean and deduplicated Arabic, English and Code tokens. Fanar Prime is a 9B parameter model continually trained on the Gemma-2 9B base model on the same 1 trillion token set. Both models are concurrently deployed and designed to address different types of prompts transparently routed through a custom-built orchestrator. The Fanar platform provides many other capabilities including a customized Islamic Retrieval Augmented Generation (RAG) system for handling religious prompts, a Recency RAG for summarizing information about current or recent events that have occurred after the pre-training data cut-off date. The platform provides additional cognitive capabilities including in-house bilingual speech recognition that supports multiple Arabic dialects, voice and image generation that is fine-tuned to better reflect regional characteristics. Finally, Fanar provides an attribution service that can be used to verify the authenticity of fact based generated content. The design, development, and implementation of Fanar was entirely undertaken at Hamad Bin Khalifa University's Qatar Computing Research Institute (QCRI) and was sponsored by Qatar's Ministry of Communications and Information Technology to enable sovereign AI technology development.
title Fanar: An Arabic-Centric Multimodal Generative AI Platform
topic Computation and Language
Artificial Intelligence
I.2.0; D.2.0
url https://arxiv.org/abs/2501.13944