Introducing TII’s Falcon H1 and Falcon Arabic
Making Advanced AI accessible and Available to Everyone, Everywhereupcoming MODELS
Powering the Future of Responsible AI
Falcon H1 and Falcon-e are the latest additions to the Falcon LLM family—designed to expand the possibilities of generative AI across industries and disciplines. Built with efficiency, ethical design, and global scalability in mind, these models reflect our commitment to making advanced AI accessible and impactful.
From smarter language understanding to enhanced real-world applications, Falcon H1 and Falcon-e pave the way for intelligent, inclusive, and future-ready solutions.
Falcon-H1
Falcon-H1: Innovative Hybrid AI Architecture Setting.....
AI innovation – and its deployment in the real world – is speeding up. Until now, users had to make a trade-off between performance and efficiency. The unique hybrid architecture of Falcon-H1 (based on Transformer and Mamba architecture) gives you...
Learn More
Falcon Arabic
Falcon Arabic: The Future of AI in the Arabic Language
Built for the Modern Standard Arabic and dialects, Falcon Arabic brings advanced language AI to the Arabic-speaking world.....
Learn MoreLatest Updates
Falcon H1 & Falcon Arabic
- Falcon-H1 redefines performance and portability, outperforming Meta’s LLaMA and Alibaba’s Qwen to enable real-world AI on everyday devices and in resource-limited settings
- Falcon Arabic: In addition to English and European-origin languages, Falcon now integrates Arabic - expanding its reach across the Arabic-speaking world as the region’s best- performing Arabic AI model
Falcon 3
- The latest updates to the Falcon 3 series introduce exceptional multimodal functionalities, enabling the models to process not only text but also images, and – for the first time in the Falcon series – video and audio data.
- Falcon 3 can run on light infrastructures, even laptops, without sacrificing performance
- The Falcon 3 ecosystem contains four scalable models tailored for diverse applications
- Falcon 3 supports several languages and is optimized for resource efficiency
- This latest iteration of TII’s open-source large language model series aims to democratize access to AI
Introducing Falcon Mamba 7B
- Falcon Mamba 7B is the first open source released State Space Language Model (SSLM), a new revolutionary architecture for Falcon models
- Falcon Mamba 7B is the no. 1 globally performing open source SSLM in the world, as independently verified by Hugging Face
- SSLMs have a low memory cost and don’t require additional memory to generate arbitrary long blocks of text
- Falcon Mamba 7B also outperforms traditional transformer architecture models such as Meta’s Llama 3.1 8B and Mistral’s 7B
- New model reflects the innovation and pioneering approach of Abu Dhabi in AI research and development
Falcon 2
Open-Source, Multilingual, and Multimodal – and is only AI Model with Vision-to-Language Capabilities. New Falcon 2 11B Outperforms Meta’s Llama 3 8B, and Performs on par with leading Google Gemma 7B Model, as Independently Verified by Hugging Face Leaderboard. Next up, we’re looking to add 'Mixture of Experts' to enhance Falcon 2’s capabilities even further.
read article
UAE’s Technology Innovation Institute Launches ‘Falcon Foundation’ to Champion Open-sourcing of Generative AI Models
The Technology Innovation Institute (TII), a leading global scientific research center and the applied research pillar of Abu Dhabi’s Advanced Technology Research Council (ATRC), today announced the launch of the ‘Falcon Foundation’. The non-profit Falcon Foundation is dedicated to advancing the development of open-source Generative AI models and to building sustainable ecosystems around open-source projects that accelerate technology development...
read article
Falcon 180B
Falcon 180B is a super-powerful language model with 180 billion parameters, trained on 3.5 trillion tokens. It's currently at the top of the Hugging Face Leaderboard for pre-trained Open Large Language Models and is available for both research and commercial use..
read article
Falcon 40B
Trained on 40 billion parameters and one trillion tokens, Falcon 40B was the first home-grown open source, large language model (LLM) to be released with weights for research and commercial use.
read articleCall for Proposals
AI71, a pioneering AI company launched by Abu Dhabi's Advanced Technology Research Council (ATRC) and VentureOne, announced a global Call for Proposals, inviting developers to submit exceptional ideas and use cases that they would like to bring to life with Falcon 10B. The best proposals will be awarded access to the necessary compute to work on their idea.
Falcon LLM Hackfest:
Wings of Innovation Challenge
Join the Technology Innovation Institute (TII) in the Falcon LLM Hackfest: Wings of Innovation Challenge, an exciting opportunity to showcase your skills and shape the future of AI with Falcon 10B!
Falcon LLMs and The Future
Initiative to Convene Leading University Researchers and Global AI Practitioners to Enhance TII’s Falcon LLM Series and Champion the Open-source Movement
The world’s leading university researchers and global AI practitioners can submit their proposals for how best to use Falcon for focused problem-solving challenges. The best submissions will be rewarded with compute and the opportunity to bring their ideas to life.
Our Research
Making Generative AI Models Accessible
Generative AI models are enabling us to create innovative pathways to an exciting future of possibilities – where the only limits are of the imagination.
Our team of AI researchers, scientists, and engineers are collaborating to achieve innovative outcomes – from AI theory to AI technologies, and are sharing their efforts with the wider AI community.
Datasets
At the heart of our cutting-edge Language Models (LLMs) lies a robust and diverse collection of datasets. Datasets are the lifeblood of our LLM models and fortify it with a wealth of linguistic knowledge and contextual understanding.
learn moreFalcon
Falcon LLM is a generative large language model (LLM) that helps advance applications and use cases to future-proof our world. Today the Falcon Arabic, Falcon Edge, Falcon H1, Falcon 3, Falcon Mamba 7B, Falcon 2, 180B, 40B, 7.5B, and 1.3B parameter AI models, as well as our high-quality REFINEDWEB dataset, form a suite of offerings.
learn more
Open Models
Why Open Source and Open Access?
As a key technology enabler, we firmly believe that innovation should be allowed to flourish. That is why we decided to open source or open access all our Falcon models. We launched our latest Falcon 180B LLM as an open access model for research and commercial use.
Together, we can mobilize and fast-track breakthrough solutions worldwide - catalyzing innovation for outsized impact.
Licensing and Security
Ensuring Compliance and Safeguarding Your Data
Falcon Arabic: Built for the Modern Standard Arabic and dialects, Falcon Arabic brings advanced language AI to the Arabic-speaking world.
Falcon H1: AI innovation – and its deployment in the real world – is speeding up. Until now, users had to make a trade-off between performance and efficiency. The unique hybrid architecture of Falcon-H1 (based on Transformer and Mamba architecture) gives you a family of AI models that delivers the best of both worlds – combining strong general-purpose comprehension with highly efficient processing. Thanks to an updated version of Maximal Update Parametrization it is possible to train the models reliably at very large scales; in fact, it makes the process of increasing model size much safer and more efficient.
Read more about our technical blog
Falcon Edge: Falcon Edge redefines what’s possible with small-scale language models. As AI moves beyond the cloud and into real-world devices, Falcon-e is built to work where infrastructure is limited, running fast and efficiently on CPUs, not GPUs.
Falcon 3:The latest updates to the Falcon 3 series introduce exceptional multimodal functionalities, enabling the models to process not only text but also images, and – for the first time in the Falcon series – video and audio data. Falcon 3 can run on light infrastructures, even laptops, without sacrificing performance. The Falcon 3 ecosystem contains four scalable models tailored for diverse applications. Falcon 3 supports several languages and is optimized for resource efficiency. This latest iteration of TII’s open-source large language model series aims to democratize access to AI.
The download of Falcon 3 is subject to our Terms & Conditions and Acceptable Use Policy
Falcon Mamba 7B: Falcon Mamba 7B is the first open source released State Space Language Model (SSLM), a new revolutionary architecture for Falcon models. Falcon Mamba 7B is the no. 1 globally performing open source SSLM in the world, as independently verified by Hugging Face. SSLMs have a low memory cost and don’t require additional memory to generate arbitrary long blocks of text. Falcon Mamba 7B also outperforms traditional transformer architecture models such as Meta’s Llama 3.1 8B and Mistral’s 7B. New model reflects the innovation and pioneering approach of Abu Dhabi in AI research and development.
The download of Falcon Mamba 7B is subject to our Terms & Conditions and Acceptable Use Policy
Falcon 2: Open-Source, Multilingual, and Multimodal – and is only AI Model with Vision-to-Language Capabilities. New Falcon 2 11B Outperforms Meta’s Llama 3 8B, and Performs on par with leading Google Gemma 7B Model, as Independently Verified by Hugging Face Leaderboard. Next up, we’re looking to add 'Mixture of Experts' to enhance Falcon 2’s capabilities even further.
The download of Falcon 2 is subject to our Terms & Conditions and Acceptable Use Policy
Falcon 40B: Falcon 40B is made available under the permissive Apache 2.0 software license for researchers and commercial users alike. End-users have access to any patent covered by the software in question. Apache 2.0 ensures the security and availability of safe and powerful open-source software and establishes an effective governance model.
Falcon 180B: Falcon-180B is accessible to developers through a royalty-free license, based on Apache 2.0. The licence includes restrictions on illegal or harmful use, and requires those intending to provide hosted access to the model to seek additional consent from TII.
The model is free of charge to download, use and integrate into applications and end user products.
Those hosting providers wishing to provide shared instances of the model or its derivatives as a managed service (for inference or fine tuning) are not covered by the proposed license, and are invited to enter into a separate license arrangement with TII.
The download of Falcon 180B is subject to our Terms & Conditions and Acceptable Use Policy
Falcon LLM
Pioneering the Next Generation of
Language Models
Falcon LLMs and The Future
Pioneering Language Models Shaping Tomorrow's Possibilities
Generative AI models like Falcon Arabic, Falcon Edge, Falcon H1, Falcon 3, Falcon Mamba 7B, Falcon 2, Falcon 40B and Falcon 180B are illuminating pathways to the future. As newer and more powerful models continue to emerge every day, they will remain the UAE’s iconic trailblazers.
We work tirelessly to ensure the transparency and accountability of our AI systems to protect privacy and data security, as well as establish mechanisms for addressing AI’s societal impact. By sharing our advancements, we facilitate breakthroughs in various domains, including healthcare, finance, education, and beyond. Together, we can leverage collective intelligence to solve complex challenges and drive innovation.
Falcon LLM - Join us in innovating for a better world.
Frequently Asked Questions
Yes.
The license allows companies and developers to integrate Falcon 180B into their applications and services, and host it in their own or leased infrastructure.
Yes.
The license allows companies and developers to build and monetize new applications on top of Falcon 180B and host it in their own or leased infrastructure.
Yes.
The license allows companies and developers to use Falcon 180B within their internal tools and applications, and host it in their own or leased infrastructure.
Yes.
The license doesn’t not prevent hosting providers from leasing hosting capacity / compute power that is used by developers or companies to host their dedicated instance of Falcon 180B.
The license only restricts the provisioning of shared instances offered as an inference / finetuning API (subject to another license)
No.
The license specifically restricts hosting providers from monetizing direct access to shared instances of Falcon 180B and its fine-tunings.
This would require a specific license to be in place.
Remember, these FAQs are based on the provided terms and conditions and aim to simplify the understanding of Falcon 180B's usage. For comprehensive information, always refer to the full License Agreement.