OpenAI Outlines European Expansion and Platform Strategy at Berlin DevDay 2026
At DevDay in Berlin, OpenAI detailed European rollout timelines for Dots, massive Codex growth in Germany, and its strategy to lock developers into managed agent infrastructure.
At OpenAI’s DevDay Exchange 2026 event in Berlin, the company largely reflected the product updates previously revealed in San Francisco. However, the European gathering highlighted critical updates regarding regional product availability, usage patterns for coding tools, and OpenAI’s broader strategy to establish an integrated application ecosystem.
Romain Huet, OpenAI’s head of developer experience, addressed the European availability of the company’s AI assistant, Dots. Huet stated that Dots is expected to expand across Europe within a few weeks following discussions with local regulators. Currently, Personal Pro tier subscribers across the European Economic Area, Switzerland, and the UK remain unable to access the service, though select enterprise accounts have already been granted access.
Alongside the expansion plans for Dots, OpenAI shared metrics on tool adoption in Europe. In Germany, weekly active users of Codex have expanded twentyfold since January 2026, building on a sevenfold increase reported in June. While absolute user metrics were not disclosed, OpenAI noted that nearly 20 percent of desktop users now employ Codex skills for tasks such as managing spreadsheets, generating presentations, and editing documents, demonstrating that the programming utility is increasingly utilized for everyday office tasks.
Technical priorities and resource allocation
During a Q&A session, attendees questioned why OpenAI has not released a model capable of processing multi-frame video sequences, a feature offered by competitor models like Google's Gemini. OpenAI management responded that demand from users for sequential video analysis has been minimal, leading the company to limit its current focus to single-frame image processing.
This rationale extends to OpenAI’s decisions regarding dedicated video tools. The company reiterated that its Sora video generator was turned off in March due to intense computational demands, noting that customer interest was heavily skewed toward coding engines and autonomous agents. However, management indicated that a revival of Sora in the future has not been entirely ruled out.
The rapid expansion of autonomous agents—which execute code on isolated virtual environments—has driven up requirements for traditional compute power. To mitigate CPU overhead resulting from rapid AI-generated code expansion, OpenAI maintains a specialized engineering team that conducts weekly reviews of internal codebases to optimize performance.
Platform ecosystem and vendor integration
Beyond specific feature announcements, the event showcased OpenAI's efforts to build a sticky platform rather than functioning solely as a model vendor. Through its Agents API, OpenAI hosts and manages the core infrastructure required to run automated workflows. Coupled with support for web apps and plugins, the platform encourages creators to deploy services directly to ChatGPT’s reported 1.2 billion weekly active users.
This strategy closely mirrors historical mobile app store models, where the platform provider handles distribution and hosting while developers supply application logic. Although OpenAI maintains the underlying execution engine, Codex Harness, as open-source code, developers relying heavily on managed backend tools and platform-specific entry points may face high switching costs if they attempt to migrate later.
What it means for developers
For software engineers and product teams, OpenAI's direction signals a clear pivot toward managed agent infrastructure and integrated distribution channels. While building directly on top of the Agents API lowers technical friction and provides direct access to a massive user base, it also deepens long-term dependencies on OpenAI's proprietary cloud services.
Developers must weigh these platform advantages against potential vendor lock-in. For teams evaluating multiple architectures or seeking flexibility, developers can try top AI models cheaply through one API at https://apixoai.online to compare different model capabilities before committing to a single infrastructure provider.
Additionally, the strong shift toward general office automation via Codex suggests growing opportunities for specialized workflows. As AI-generated code increases baseline CPU demands, developers will also need to place greater emphasis on runtime efficiency and infrastructure management when deploying autonomous agents.
Source: OpenAI teases Dots for Europe, but its Berlin DevDay was really about building a platform — The Decoder. Written by the Apixo team from that report.
One key for Claude, GPT, GLM, DeepSeek and more. Pay per token with crypto.
Get your API key
