Cerebras Systems’ disruptive technology has shaken the tech sector over the last decade. Just last week, the company announced it had closed on $1 billion in new capital, increasing its valuation to a staggering $23 billion. This historic investment is the most recent indication that the tide has turned for the company. It is symptomatic of its increasing reach and power during this booming age of artificial intelligence (AI). Cerebras recently announced plans to make its debut on the public market with an initial public offering (IPO), confirming its stronghold in the burgeoning market.
As we wrote earlier this month, OpenAI released a new agentic coding model, Codex. The company simultaneously announced the release of a lightweight version, GPT-5.3-Codex-Spark, which is optimized for faster inference. This more compact edition seeks to be a true “daily productivity driver,” helping its users with fast iteration and in-context collaboration.
The original Codex model is best suited for longer-form tasks that involve more complex reasoning and task orchestration. GPT-5.3-Codex-Spark has a heavy focus on promoting fast iteration and novel interaction patterns. OpenAI’s CEO, Sam Altman, hinted at the launch of this model through a recent tweet, stating, “We have a special thing launching to Codex users on the Pro plan later today.”
Cerebras is working to incorporate its technology into OpenAI’s compute solutions, which should increase the speed at which AI can respond. Sean Lie, the CTO and co-founder of Cerebras, shared his excitement over this partnership. He stated, “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.”
This integration is seen as the first step toward a Codex that operates in two complementary modes: one for rapid iteration and another for long-running tasks requiring deeper reasoning.
