OpenAI Combines Models, Cloud Tools, and Marketplace in Its Strategy

OpenAI has implemented a new distribution model: professionals now have the option to use a portion of their commitments to acquire products offered by partners, while integrations allow them to use their own tokens on other services. At the same time, the company is introducing cloud tools for developers, presenting a new model, and offering Ultrafast access. According to OpenAI, ChatGPT has over 1.2 billion weekly users.

Redirected Spending and Interoperable Tokens for Professional Clients

Sam Altman clarified that professional clients now have the ability to allocate a portion of their existing commitments with OpenAI to purchase products marketed by partners. Users can also link ChatGPT to other AI solutions to utilize some of their own allocated tokens. According to OpenAI, more than 1.2 billion people use ChatGPT each week. OpenAI is thus reinforcing its role as an intermediary between businesses, while other players seek to connect to its platform to attract users or integrate their existing bases with its technology. Sam Altman views cloud computing providers as a relevant example to support the ecosystem and hopes OpenAI will become a central hub to drive demand and connect clients. He states his intention to widely deploy the technology, attract people, revenue, and clients, and contribute to making the ecosystem as robust and rich as possible.

Building Tools: Codex AI in the Cloud and Two New APIs

OpenAI has added Codex AI in the cloud, allowing developers to use it on various devices. The company also introduced an API for creating agents and another dedicated to rapid decision-making, which received strong applause from the audience during the conference. Sam Altman aims to provide developers with the same tools that OpenAI uses internally and make the platform a one-stop shop for application building. The Jev model from the startup TypeSafe AI offers a similar service to the rapid decision-making API.

Model Offerings: GPT-6.1 Sol, Ultrafast Access, and Research Priority

Sam Altman announced the GPT-6.1 Sol model, presented as an affordable alternative to the Astra model. He also unveiled Ultrafast, a mode of access to OpenAI's models that offers enhanced performance but at a higher cost. The stated goal is to provide the best combination of intelligence and capability at every price point across the market. OpenAI still dedicates the majority of its computing power to research. The company operates in a highly competitive environment, with Anthropic, Meta, Google, and SpaceXAI, while low-cost Chinese offerings pose an additional threat. OpenAI faces numerous rivals as it began as a research lab, and the strategy is part of a long-term struggle against these players.

The Three-Pronged Plan and the Context of the Announcement in San Francisco

During the DevDay conference in San Francisco, Sam Altman, CEO of OpenAI, presented a three-pronged plan structured around models, building for developers, and distribution in the form of a marketplace. He specified his desire to provide the best AI models, facilitate application creation, and enable professional clients to connect with each other. The presentation was accompanied by updates for ChatGPT and other products, as ChatGPT claims 1.2 billion weekly users. The context highlights the pressure on OpenAI and Anthropic to support high valuations with ambitious business plans, against a backdrop of strong growth since the emergence of ChatGPT.