무슨 일이 일어났나요?
The Presidency of Defense Industries (SSB) announced the launch of EVREN, a national artificial‑intelligence platform aimed at Turkey’s defense sector. The platform aggregates datasets, AI models and a high‑performance GPU cluster on a single portal (evren.ssyz.org.tr) and requires e‑Government (e‑Devlet) authentication for access. Rather than charging fees, EVREN uses a credit model: users earn credits by uploading or labeling data and by sharing trained models, which can then be spent on model training and large‑language‑model . The service includes end‑to‑end computer‑vision tools (object detection, segmentation, ) and an inference layer for 11 open‑weight LLMs. API calls made before 1 Nov 2026 are exempt from credit deductions. Within a short period, the platform reported 7,500 active users. Future plans include adding domestically developed LLMs, image and audio models, and expanding GPU capacity via a distributed management approach.
On Sunday, the SSB released a statement confirming the operational launch of EVREN, a national AI platform designed for defense‑related use cases. The platform aggregates data, AI models and a high‑performance GPU pool under a single web portal, accessible via the Turkish e‑Government authentication system.
EVREN’s access model is contribution‑based: users earn credits by contributing datasets, labeling data, or sharing trained models. These credits can be spent on GPU‑intensive tasks such as model training and with 11 open‑weight large language models. The platform also offers computer‑vision pipelines for object detection, segmentation and .
To encourage early adoption, the SSB announced that API calls made before 1 Nov 2026 will not deduct from users’ credit balances. Within a short launch window, the platform reported 7,500 active users, indicating rapid uptake among defense firms, technology companies, academics and students.
Future development plans include integrating domestically created LLMs, image and audio models, and expanding the GPU infrastructure through a distributed management approach.
왜 중요한가요?
EVREN represents a strategic move by Turkey to build a sovereign AI infrastructure for its defense industry, reducing reliance on foreign cloud services that could expose sensitive data. By keeping data processing on domestic GPU hardware, the platform addresses national security concerns while fostering a collaborative ecosystem where contributors are incentivized to share resources. The credit‑based model lowers entry barriers for smaller firms, academia and students, potentially accelerating AI talent development and innovation within the country. Moreover, the platform’s open‑weight LLMs and computer‑vision capabilities could speed up prototype development for defense applications, from autonomous systems to intelligence analysis. However, the lack of publicly disclosed pricing, model performance benchmarks, and external validation means the platform’s practical impact remains to be seen.
Sovereign AI infrastructure mitigates the risk of sensitive defense data being processed on foreign cloud services, aligning with national security priorities.
The credit‑based contribution system lowers financial barriers, potentially democratizing access to high‑end AI resources for smaller enterprises and research groups.
By centralizing data and models, EVREN could accelerate the development of AI‑driven defense technologies, such as autonomous vehicles, surveillance analytics and decision‑support tools.
The platform’s reliance on open‑weight LLMs and domestic model development may foster a home‑grown AI talent pipeline, but the absence of independent performance evaluations leaves the efficacy of these models uncertain.
대화형 메커니즘: 실제로 작동하는 방식
이 개발의 이면에 있는 기본 기술을 대화식으로 살펴보세요.
crm_get_transaction(id='4092').Which component of an AI application is the machine-learning model itself?
다음에 무엇을 볼 것인가
Key indicators to monitor include: (1) adoption rates beyond the initial 7,500 users, especially among defense contractors and research institutions; (2) the rollout of domestically built LLMs and whether they meet performance standards comparable to foreign models; (3) any policy changes to the credit system or extensions of the free‑API period; and (4) potential export controls or international reactions to a sovereign defense AI platform.
User growth beyond the initial 7,500, especially among established defense contractors, will indicate the platform’s commercial viability.
The performance and adoption of domestically built LLMs will reveal whether Turkey can achieve parity with international AI offerings.
Any revisions to the credit system or extensions of the free‑API period could affect long‑term sustainability and user incentives.
International response, including potential export controls or diplomatic concerns, may shape the platform’s future scope and collaborations.