Developed under the leadership of the Presidency of the Republic of Turkey’s Defense Industry Directorate, the domestic and national artificial intelligence platform EVREN has been officially launched with the aim of strengthening Turkey’s information and technological sovereignty. Targeting a wide audience from the defense industry to technology startups, from academics to students, the platform brings together data sets, advanced models, and high-performance GPU infrastructure in a single center.

Data Sovereignty Guarantee: All Processes on Domestic Infrastructure

All request, response, and usage information on EVREN is processed directly on the high-performance GPU infrastructure within Turkey’s borders. Thanks to this architectural approach, the leakage or transfer of critical and sensitive information to foreign-sourced cloud services is completely prevented. Users can securely log in to the platform via the address evren.ssyz.org.tr by completing the e-Government identity verification steps.

Access to GPU Power with Contribution-Based Credit Model

Unlike classic subscription or direct-price models, the platform builds a community-focused ecosystem. Users earn credits by uploading datasets, performing data labeling, or sharing models they have previously trained.

These earned credits are used to train new models and perform large language model inferences in the platform’s high-processing-capacity GPU pool. This mechanism positions developers not only as consumers but also as personal developers of the system.

11 Open Weighted Model and Computer Vision Support

EVREN provides end-to-end data preparation and model training capabilities in fundamental computer vision areas such as object detection, segmentation, and classification.

The platform also features an advanced inference layer encompassing 11 different open-loaded large language models. Thanks to real-time response reinforcement, automatic model guidance, and a modern API architecture, developers can quickly integrate their existing software into the EVREN infrastructure. To accelerate ecosystem adoption, all API invitations made until November 1, 2026, are offered unlimitedly without deduction from credit balances.

Target: Native Model and Distributed GPU Ecosystem

Having surpassed 7,500 active users in a short time, EVREN aims to directly contribute to the goals of the National Artificial Intelligence Action Plan.

In the coming period, it is planned to integrate native large language, image, and voice models into the system, while further scaling of computing capacity with distributed GPU management is targeted. The platform; By bringing together academia, the public and private sectors on a common innovation platform, it is taking on the main backbone role in Turkey’s artificial intelligence initiative.