Jais 2 Params: 70B | ALLaM 34B: Live | Falcon-H1 OALL: 75.36% | MENA AI Funding: $2.1B H1 | HUMAIN Infra: $77B | Arabic Speakers: 400M+ | OALL Models: 700+ | Saudi AI Year: 2026 | Jais 2 Params: 70B | ALLaM 34B: Live | Falcon-H1 OALL: 75.36% | MENA AI Funding: $2.1B H1 | HUMAIN Infra: $77B | Arabic Speakers: 400M+ | OALL Models: 700+ | Saudi AI Year: 2026 |

MENA AI Ecosystem Terminology — Organizations and Initiatives Glossary

Glossary of MENA AI ecosystem terminology — organizations, initiatives, strategies, and the institutional landscape of Arabic artificial intelligence.

Advertisement

The Middle East and North Africa AI ecosystem has evolved from isolated research projects into a multi-billion-dollar institutional landscape spanning sovereign wealth funds, national AI strategies, research universities, and an accelerating startup scene. This glossary defines the organizations, initiatives, programs, and infrastructure that constitute the Arabic AI ecosystem as of 2026, with cross-references to detailed profiles and analysis throughout this site. The UAE and Saudi Arabia together account for 87 percent of MENA AI venture capital investment, having attracted $519 million and $235 million respectively in a single year from 322 deals.


A

AI Sovereignty — The strategic objective of a nation developing indigenous AI capabilities rather than depending on foreign technology providers. Both Saudi Arabia and the UAE have adopted AI sovereignty as a core policy goal, driving investments in domestically developed large language models (Jais in the UAE, ALLaM in Saudi Arabia), national data centers, and local talent development. AI sovereignty for Arabic-speaking nations has a linguistic dimension absent from most Western discussions — dependence on English-centric AI models means Arabic language and cultural knowledge is mediated through foreign systems that may not represent Arab values, dialects, or communicative norms accurately.

ALLaM — Arabic Large Language model developed initially by NCAI at SDAIA and now managed by HUMAIN. Available in 7B, 13B, 34B, and 70B parameter sizes. The 13B instruct version was trained on 3 trillion tokens of Arabic and English data. ALLaM’s training data included input from 16 public entities, 300 Arabic books, 400 subject matter experts, and over 1 million test prompts. Available on IBM watsonx (May 2024), Microsoft Azure (September 2024), and Hugging Face (early 2025). The ALLaM Challenge is a developer competition with SAR 1 million in prizes for Arabic AI applications. Ranked as the world’s most advanced Arabic LLM built in the Arab world by Cohere on the MMLU benchmark.

ALLaM Challenge — Developer competition organized by SDAIA/HUMAIN offering SAR 1 million in prizes for innovative Arabic AI applications built on the ALLaM foundation model. The challenge drives ecosystem development by encouraging Arabic developers to build commercial applications on domestically developed AI infrastructure rather than defaulting to GPT-4 or other Western models.

ASPIRE — Arabic name for Saudi Arabia’s National Strategy for Data and AI (NSDAI), managed by SDAIA. Built on six strategic pillars with 66 specific targets across three phases: national urgencies by 2025, competitive advantage by 2030, and global leadership post-2030. Targets include training 20,000 AI specialists, incubating 300 AI startups, and attracting over $20 billion in AI investment. Saudi Arabia ranked 14th in the 2025 Global AI Index and first globally in public sector AI adoption.

C

CAMeL Lab — Computational Approaches to Modeling Language Lab at NYU Abu Dhabi, established September 2014 under the direction of Dr. Nizar Habash. One of the most productive Arabic NLP research groups globally, responsible for CAMeL Tools (Python suite for Arabic NLP), MADAMIRA (morphological tagger), CALIMA Star (morphological analyzer), YAMAMA (fast multi-dialect analyzer), and CAMeL Parser (dependency parser). The lab has also produced critical Arabic corpora including MADAR (parallel sentences in 25 city dialects plus English, French, and MSA), GUMAR (100 million word Gulf Arabic corpus), CAMeLTB (188,000 word dependency treebank spanning pre-Islamic poetry to social media), QALB (2 million word manually corrected corpus), and SAMER (26,000 lemma readability lexicon for MSA).

Cerebras — US-based AI hardware company that builds wafer-scale computing systems, where an entire silicon wafer serves as a single processor rather than being diced into individual chips. Partner in Jais LLM development with G42 and MBZUAI. Cerebras provided the hardware architecture for the Condor Galaxy supercomputer that trained Jais models. The partnership demonstrates how MENA AI development leverages international hardware partnerships while maintaining sovereign control over model training, data, and deployment.

Condor Galaxy — Multi-exaFLOP AI supercomputer built by G42 and Cerebras Systems. The primary training infrastructure for Jais models, providing the massive compute required to train Arabic-first large language models from scratch. Condor Galaxy 1 (CG-1) was the initial system, with subsequent expansions planned to support larger model training runs. The system uses Cerebras’ wafer-scale engine architecture, offering higher throughput for training workloads than traditional GPU clusters.

E

Egypt AI Fund — A $300 million investment fund established through a partnership between Egypt and Tsinghua Unigroup to support AI development in Egypt and the broader North African region. Part of Egypt’s strategy to position itself as an AI hub for Arabic-speaking Africa, leveraging its large population (over 100 million) and established tech talent pipeline.

F

Fanar — Arabic-centric multimodal generative AI platform developed by QCRI (Qatar Computing Research Institute) in 2025. Represents Qatar’s entry into sovereign Arabic AI, complementing the UAE’s Jais and Saudi Arabia’s ALLaM. Fanar focuses on multimodal capabilities including text, image, and potentially audio processing in Arabic.

Falcon — Family of large language models developed by TII (Technology Innovation Institute) in Abu Dhabi. The series spans Falcon 1 (2023), Falcon 2 (Spring 2024), and Falcon 3 (December 2024). Falcon Arabic (7B, May 2025) was the first Arabic-dedicated model from TII, trained on 600 billion tokens of Arabic, multilingual, and technical data. Falcon-H1 Arabic (January 2026) introduced a hybrid Mamba-Transformer architecture in 3B, 7B, and 34B sizes with 256K token context windows. Falcon-H1 34B achieved the highest score (75.36 percent) on the Open Arabic LLM Leaderboard. Licensed under Apache 2.0-based TII Falcon License.

G

G42 (Group 42) — UAE-based AI and cloud computing company headquartered in Abu Dhabi. Developer of the Jais LLM series in partnership with MBZUAI and Cerebras Systems. Received a landmark $2.3 billion investment from Microsoft in 2024, signaling Western technology companies’ strategic interest in the Gulf AI ecosystem. G42’s portfolio spans the full AI value chain from foundational infrastructure (Condor Galaxy supercomputer) through model development (Jais) to commercial cloud computing and computer vision applications. G42 also partners with OpenAI on the Stargate UAE project, a planned 1 GW AI computing cluster in Abu Dhabi.

GAIA Accelerator — Regional AI accelerator program with a $1 billion budget, established through collaboration between SDAIA, New Native, and NTDP (National Technology Development Program). GAIA aims to accelerate Arabic AI startups from ideation through commercialization, providing funding, compute resources, mentorship, and market access. The accelerator reflects Saudi Arabia’s strategy of developing a domestic AI startup ecosystem rather than relying solely on imported technology.

H

Hub71 — Abu Dhabi’s global tech ecosystem, providing incentives, funding, and community infrastructure for technology startups including AI companies. Hub71 has become a landing point for international AI startups entering the MENA market, offering subsidized office space, housing, and healthcare alongside access to UAE government contracts and G42 cloud resources.

HUMAIN — Saudi Arabia’s national AI company, launched May 12, 2025 by Crown Prince Mohammed bin Salman. Backed by the Public Investment Fund (PIF). Ambition is to become the third-largest AI provider globally, behind only the United States and China. HUMAIN manages ALLaM development and operates an infrastructure program comprising 11 data centers across two campuses at 200 MW each, ramping at 50 MW per quarter from Q4 2025. Targets 1.9 GW by 2030 and 6 GW by 2034 at an estimated total cost of $77 billion. HUMAIN has signed $23 billion in deals since May 2025, with partners including xAI (500 MW data center), Adobe (first global tenant), NVIDIA, AMD, and AWS. Plans a $10 billion venture fund for AI startups. HUMAIN Chat is the national Arabic AI chatbot featuring real-time web search, Arabic speech input supporting multiple dialects, bilingual Arabic-English switching, and Saudi PDPL compliance. See the full HUMAIN profile and data center analysis.

J

Jais — World’s most advanced Arabic open-weight large language model, developed by G42 Inception, MBZUAI, and Cerebras Systems in the UAE. Named after Jebel Jais, the highest mountain in the UAE. Versions include Jais-13B (August 2023, 116B Arabic + 279B English training tokens), Jais-30B (November 2023), the Jais Family 2024 (20 open-source models from 590M to 70B parameters — the largest single release in MENA), and Jais 2 (December 2025, 70B parameters, 600B+ Arabic tokens). Capabilities include MSA and 17 regional dialects, Arabizi recognition, code-switching, Arabic poetry, and technical/creative content. Available on Hugging Face and through JaisChat.ai. Designed to serve 400 million+ Arabic speakers worldwide.

M

MAGNiTT — The leading startup data platform for MENA, providing funding data, company profiles, and ecosystem analytics. MAGNiTT data shows that AI’s share of total MENA venture capital reached 22 percent ($858 million) in 2025, with H1 2025 total reaching $2.1 billion — a 134 percent year-over-year increase. Saudi Arabia alone recorded $860 million in H1 2025 across 114 deals, a 116 percent year-over-year increase.

MBZUAI (Mohamed bin Zayed University of Artificial Intelligence) — The world’s first graduate-level university dedicated entirely to AI, based in Abu Dhabi. Co-developer of Jais with G42 and Cerebras. MBZUAI provides the academic research foundation for the UAE’s AI ambitions, training PhD and Master’s students in machine learning, NLP, computer vision, and robotics. The university’s research focus on Arabic NLP provides the linguistic expertise that complements G42’s engineering capabilities in Jais development. See the full MBZUAI profile.

N

NCAI (National Centre for Artificial Intelligence) — The SDAIA unit that originally developed ALLaM. NCAI’s research team built the initial ALLaM models before the program transitioned to HUMAIN in 2025. NCAI continues to support Saudi Arabia’s broader AI research agenda beyond large language models.

NSDAI — National Strategy for Data and AI, Saudi Arabia’s comprehensive national AI strategy managed by SDAIA. Also known by its Arabic name ASPIRE. The strategy sets ambitious targets across three time horizons with six pillars covering government AI adoption, private sector development, talent creation, data infrastructure, ethics and governance, and international partnership.

O

OALL (Open Arabic LLM Leaderboard) — Benchmark platform hosted on Hugging Face, co-developed by 2A2I, TII, and Hugging Face. Launched May 2024, the leaderboard has received over 700 model submissions from more than 180 organizations. Version 2 removed machine-translated evaluation tasks entirely, replacing them with native Arabic benchmarks including ArabicMMLU, ALRAGE, AraTrust, and MadinahQA. The OALL tracks performance across LLM performance, multimodality/vision, embedding, retrieval, RAG generation, speech-to-text, and OCR. See the full OALL analysis.

P

PIF (Public Investment Fund) — Saudi Arabia’s sovereign wealth fund and one of the world’s largest, with over $900 billion in assets under management. Parent entity of HUMAIN. PIF’s backing gives HUMAIN the financial resources to execute its $77 billion data center infrastructure program and $10 billion startup venture fund. PIF also invests directly in international AI companies, creating strategic partnerships that benefit Saudi Arabia’s domestic AI ecosystem.

Polynome Group — A $100 million AI-focused investment fund launched in Q1 2025, targeting MENA AI startups. Part of the expanding MENA AI venture capital landscape alongside Smpl Fund I ($10 million) and HUMAIN’s planned $10 billion venture fund. These dedicated AI funds complement generalist VC firms that have increased their AI allocation in response to the sector’s growth.

Project Transcendence — Saudi Arabia’s $100 billion AI initiative announced late 2024, focusing on world-class data centers, AI startup incubation, international talent recruitment, and technology partnerships. Project Transcendence represents the most ambitious single-country AI investment outside the United States and China. The initiative encompasses HUMAIN’s infrastructure program, the GAIA accelerator, and strategic partnerships with global technology companies.

Q

QCRI (Qatar Computing Research Institute) — Qatar’s primary AI research institution, part of Hamad Bin Khalifa University. Developed Fanar, an Arabic-centric multimodal generative AI platform launched in 2025. QCRI has a long history of Arabic NLP research predating the current LLM era, including contributions to Arabic machine translation, speech recognition, and social media analysis.

S

SDAIA (Saudi Data and Artificial Intelligence Authority) — Saudi Arabia’s national AI strategy and governance body, established August 2019. Manages the NSDAI/ASPIRE strategy built on six pillars with 66 targets. Priority sectors include education, healthcare, energy, mobility, and government. Under SDAIA’s leadership, Saudi Arabia achieved first place globally in public sector AI adoption and 14th place in the 2025 Global AI Index. SDAIA launched the GAIA accelerator ($1 billion budget with New Native and NTDP) and oversaw initial development of ALLaM before transitioning the program to HUMAIN. See the full SDAIA strategy analysis.

Stargate UAE — A planned 1 GW AI computing cluster in Abu Dhabi, a partnership between OpenAI and G42. The project reflects the UAE’s strategy of attracting international AI companies to build infrastructure within the country, ensuring that AI computing capacity is physically located in the Gulf region rather than concentrated in North American data centers.

T

TII (Technology Innovation Institute) — Abu Dhabi-based research institute responsible for the Falcon LLM series. TII operates across multiple technology domains but is best known in the AI field for producing some of the world’s most capable open-source language models. TII co-manages the OALL with 2A2I and Hugging Face. Falcon-H1 Arabic (January 2026) is TII’s most advanced Arabic model, using a hybrid Mamba-Transformer architecture that leads the OALL benchmarks. See the full TII profile.

V

Vision 2030 — Saudi Arabia’s national transformation program targeting economic diversification, social reform, and the development of a post-oil knowledge economy. AI is a central pillar of Vision 2030, with the Year of AI 2026 designation by the Saudi Cabinet reflecting AI’s strategic importance. Saudi Arabia now has 664 AI companies and recorded $9.1 billion in AI funding through 70 deals in 2025.

Y

Year of AI 2026 — Designation by the Saudi Cabinet declaring 2026 as the Year of AI, reflecting the sector’s strategic priority within Vision 2030. As of the designation, Saudi Arabia has 664 AI companies and has secured $9.1 billion in AI funding through 70 deals. The designation signals continued government-level commitment to AI development and is expected to drive further policy support, funding allocation, and international partnership development. See the full Year of AI analysis.


Advertisement
Advertisement

Institutional Access

Coming Soon