What happened
OpenAI launched the Decisions API on October 6, 2026, introducing a specialized service for rapid, automated decision-making that directly competes with TypeSafe's Jev model. According to Fortune, the new API leverages the existing GPT-6 Luna model but constrains it to return structured answers, enabling it to process multiple questions simultaneously and deliver initial decisions up to ten times faster than standard GPT-6 Luna calls. The service is priced at $0.10 per million input tokens, with no charge for output tokens, positioning it as a cost-effective alternative for high-volume tasks such as routing customer service messages or flagging content. OpenAI executives stated that the product was not on the development roadmap four weeks prior to its release, indicating a rapid strategic pivot in response to the market traction gained by TypeSafe's Jev, which launched three weeks earlier.
OpenAI released the Decisions API on October 6, 2026, just three weeks after TypeSafe launched its competing Jev model. Fortune reports that the Decisions API is built on the existing GPT-6 Luna model but is specifically constrained to return structured answers, allowing it to handle multiple questions at once and deliver the first decision faster than general-purpose calls.
The service is designed for applications that need to choose the right model, tool, or action in near real time. OpenAI claims the dedicated interface can return decisions up to ten times faster than calling GPT-6 Luna through its standard Responses API. The API accepts text and images, allowing applications to check photographs or browser screenshots, whereas Jev currently accepts only text.
Pricing for the Decisions API is listed at $0.10 per million input tokens, with no charge for output tokens. This compares to Jev's price of $0.042 per million tokens. OpenAI's Nikunj Handa stated in a podcast interview that the Decisions API was not on the development roadmap four weeks before its release, suggesting it was a rapid response to the success of Jev.
TypeSafe CEO Diogo Almeida, a former OpenAI employee, founded TypeSafe to address what he called the 'massive over-promise under-deliver issue' in AI. He developed Jev to handle routine judgments faster and more cheaply than general-purpose LLMs, using a new model architecture and training method intended to make probability estimates reflect judgment accuracy.
Why it matters
This launch marks a significant shift in how major AI providers approach routine automation, moving away from general-purpose large language models toward specialized, high-speed decision engines. By offering a dedicated interface for and routing, OpenAI addresses the cost and barriers that have historically prevented businesses from automating high-volume, low-complexity tasks. The direct competition with TypeSafe Jev highlights a growing industry consensus that 'System One' AI—fast, intuitive judgments—is a distinct and valuable market segment separate from complex reasoning. For developers, this provides a new option to reduce operational costs and improve response times for tasks like email sorting or compliance checks, potentially accelerating the adoption of AI in enterprise workflows where speed and price are critical factors.
The launch of the Decisions API signals a strategic pivot by OpenAI toward specialized, high-speed decision-making tools, acknowledging that general-purpose LLMs are often too slow and expensive for high-volume, low-complexity tasks like and routing.
By offering a dedicated interface for these tasks, OpenAI aims to capture a significant portion of the enterprise market that requires automated judgments at scale. This move could reduce the cost of AI automation for businesses, making it more feasible to deploy AI for routine operations such as customer service triage and content moderation.
The competition between OpenAI and TypeSafe highlights a broader trend in the AI industry toward specialization. As models become more capable, the focus is shifting from raw intelligence to practical efficiency, with companies developing tools that are optimized for specific use cases rather than general-purpose reasoning.
For developers, the availability of multiple options for fast decision-making allows for greater flexibility in choosing the right tool for their specific needs. This competition may drive down prices and improve performance, ultimately benefiting end-users who rely on AI-powered services.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
An agent must create a draft calendar event for Tuesday at 2 p.m. Which evidence would establish the requested result?
What to watch next
Developers should monitor the performance differences between OpenAI's Decisions API and TypeSafe Jev in real-world production environments, particularly regarding error rates and reliability for specific tasks. Additionally, the pricing dynamics between the two services may shift as OpenAI leverages its scale to adjust costs, potentially impacting the viability of specialized startups in this niche. Observers should also watch for further integration of these decision-making capabilities into broader OpenAI products, which could further compress the market for standalone classification tools.
The performance and reliability of the Decisions API in production environments will be a key factor in determining its success. Developers will need to evaluate how well it handles edge cases and maintains accuracy under high load compared to Jev and other competitors.
Pricing strategies from both OpenAI and TypeSafe will likely evolve as they compete for market share. OpenAI's scale may allow it to offer more aggressive pricing, which could pressure TypeSafe to adjust its own pricing or differentiate its product in other ways.
The integration of decision-making capabilities into broader AI platforms could further consolidate the market. If OpenAI bundles the Decisions API with other services, it may become the default choice for developers already using its ecosystem, potentially limiting the growth of specialized startups.
The adoption of 'System One' AI models in enterprise settings will be a key indicator of the market's maturity. As more businesses deploy these tools for routine tasks, the demand for fast, reliable decision-making will likely grow, driving further innovation in this space.