Skip to content
NakodaAI
← Research & Resources

Research paperInception (G42) / Cerebras Systems / Mohamed bin Zayed University of Artificial Intelligence (MBZUAI)2023-08-30

Jais and Jais-chat: Arabic-Centric Foundation and Instruction-Tuned Open Generative Large Language Models

By Neha Sengupta, Sunil Kumar Sahu, Bokang Jia, and further co-authors

The paper introducing Jais, at the time the largest Arabic-centric open large language model, built with a bilingual (Arabic/English) training and vocabulary approach rather than adapting an English-first model after the fact.

Why it matters

It is direct evidence that the UAE's AI ecosystem produces frontier-adjacent research, not only strategy documents - a joint effort between an Abu Dhabi-headquartered company (G42's Inception), a UAE university (MBZUAI) and Cerebras, openly released rather than kept proprietary.

Key takeaways

  • Built Arabic-first rather than as an English model fine-tuned on Arabic afterward, with a custom bilingual tokenizer designed for Arabic's morphology.
  • Released as open models (Jais, base) and instruction-tuned chat models (Jais-chat), with weights published rather than access-gated behind an API only.
  • Reported as the largest Arabic-centric LLM openly released at time of publication, evaluated against both Arabic and English benchmarks.
  • A successor, Jais 2, was released in 2026 by the same MBZUAI/Cerebras/Inception collaboration, extending the family to a 70B-parameter model.

UAE / MENA relevance

This is UAE-originated frontier AI research: developed by Inception, part of Abu Dhabi's G42, in partnership with MBZUAI, and openly released to the global research community - one of the clearest concrete outputs of the ecosystem the National AI Strategy set out to build.

Part of these reading paths

Related

Also worth reading

Research paperTechnology Innovation Institute (TII), Abu Dhabi2023-11-28

The Falcon Series of Open Language Models

The technical paper behind Falcon-7B, -40B and -180B - causal decoder-only language models built by Abu Dhabi's Technology Innovation Institute and, at 180B parameters and 3.5 trillion training tokens, among the largest openly documented pretraining runs at time of release.

  • Research Foundations
  • UAE & MENA AI

Research paperOpenAI2023-03-15

GPT-4 Technical Report

OpenAI's technical report on GPT-4, a large multimodal model accepting image and text input, covering its benchmark performance and a substantial section on the safety evaluation and mitigation work done before release.

  • Research Foundations
  • AI Safety & Alignment

Academic resourceGaoling School of Artificial Intelligence, Renmin University of China (with Université de Montréal)2023-03-31

A Survey of Large Language Models

A continuously updated academic survey covering the background, key findings and mainstream techniques of large language models, organised around pre-training, adaptation tuning, utilisation and capability evaluation.

  • Research Foundations