Abu Dhabi Institute Releases Fully Open-Source AI Models
Abu Dhabi's IFM has released six AI models with full transparency, including training data and code, setting a new standard for openness.
Abu Dhabi's research institute IFM has unveiled six artificial intelligence models, making them fully open-source by sharing not just the models but also the training data, code, methodologies, and intermediate checkpoints. This move, announced on Thursday, allows researchers to trace the development process and reproduce results, according to IFM founder Eric Xing.
The release, named K2 Horizon, goes beyond the 'open-weight' approach used by some Chinese developers, which offer downloadable models but little insight into their construction. It also contrasts sharply with the closed practices of companies like OpenAI and Anthropic, which neither release their models nor disclose underlying data and techniques.
Xing stated that the goal is to establish a reference point for what a truly open model release can look like. He added that the initiative aims to demonstrate to policymakers, regulators, and public-interest advocates that openness and competitive performance are not mutually exclusive.
The launch is part of the UAE's broader strategy to position itself as a global AI hub. The model family spans from a version designed for smartwatches and other constrained devices to a massive 375-billion-parameter system intended for enterprise use.