Krux
February 16, 2026
OpenAI Ditches Nvidia for New Coding Model
Published: February 16, 2026 at 4:47 AM
Updated: February 16, 2026 at 4:47 AM
100-word summary
OpenAI just launched GPT-5.3-Codex-Spark, its first production model running on Cerebras Systems chips instead of Nvidia hardware. This streamlined, lower-power variant specializes in fast, interruptible coding tasks like editing specific code sections and running targeted tests. Spark runs on Cerebras' third-generation Wafer Scale Engine, featuring hundreds of thousands of AI cores on a single wafer-scale processor optimized for low-latency workflows. Rolling out Thursday as a research preview to ChatGPT Pro subscribers, developers can swap existing OpenAI endpoints for Cerebras in just 15 seconds. This marks a strategic shift toward hardware diversification beyond Nvidia's dominance in AI compute.
What happened
OpenAI just launched GPT-5.3-Codex-Spark, its first production model running on Cerebras Systems chips instead of Nvidia hardware. This streamlined, lower-power variant specializes in fast, interruptible coding tasks like editing specific code sections and running targeted tests. Spark runs on Cerebras' third-generation Wafer Scale Engine, featuring hundreds of thousands of AI cores on a single wafer-scale processor optimized for low-latency workflows. Rolling out Thursday as a research preview to ChatGPT Pro subscribers, developers can swap existing OpenAI endpoints for Cerebras in just 15 seconds.
Why it matters
This marks a strategic shift toward hardware diversification beyond Nvidia's dominance in AI compute.