OpenAI Unveils GPT-5.3-Codex-Spark for Real-Time Coding

OpenAI has launched GPT-5.3-Codex-Spark, a new lightweight version of its AI-powered coding tool Codex, aiming to deliver rapid, real-time coding assistance for developers. Announced on Thursday, the release underscores OpenAI’s push for faster AI-driven programming support and deepens its ongoing collaboration with chipmaker Cerebras.

Unlike the original GPT-5.3-Codex model—which is tailored for complex, long-duration coding and reasoning tasks—Codex-Spark targets swift operations requiring ultra-low latency. According to OpenAI, Spark provides improved inference speed, making it suitable for use cases like rapid prototyping and real-time code collaboration. Inference refers to the stage where an AI model processes input and produces an output, such as generating or editing code in response to user prompts.

Central to Codex-Spark’s performance is Cerebras’ Wafer Scale Engine 3, the third generation of the company’s AI megachip. This hardware features 4 trillion transistors on a single silicon wafer, enabling significant acceleration of AI computations. The partnership between OpenAI and Cerebras, first announced last month, is valued at over $10 billion and seeks to improve the response times of AI services through direct integration of advanced chips into OpenAI’s data centers.

“Integrating Cerebras into our mix of compute solutions is all about making our AI respond much faster,” OpenAI previously commented about the deal. Spark is now presented as the tangible outcome of this collaboration, marking a new era in AI-assisted coding tools that prioritize speed and interactivity for users.

The new Codex-Spark tool is currently available as a research preview and is limited to ChatGPT Pro subscribers using the Codex app. This phased rollout allows OpenAI to gather feedback and refine performance ahead of broader release. Prior to the announcement, OpenAI CEO Sam Altman signaled its arrival on social media, describing the launch as “something special” for Codex users on the Pro plan.

Cerebras, founded more than a decade ago, has become a prominent provider of AI hardware. The company recently secured $1 billion in new funding, reaching a valuation of $23 billion, and announced plans to go public. Sean Lie, Cerebras’ Chief Technology Officer and co-founder, highlighted the significance of the new tool: “What excites us most about GPT-5.3-Codex-Spark is partnering with OpenAI and the developer community to discover what fast inference makes possible — new interaction patterns, new use cases, and a fundamentally different model experience.”

OpenAI’s statement frames Codex-Spark as the first step toward a dual-mode Codex system: enabling both quick, real-time collaboration and more extensive, reasoning-intensive automation. The introduction of ultra-low latency tools is expected to impact developer workflows, allowing for faster iteration and potentially opening up new approaches to AI-driven software engineering.

Reference: dataconomy.com