← Back to Briefing
OpenAI Unveils Ultra-Fast AI Coding Model Powered by Cerebras Chips
Importance: 98/1008 Sources
Why It Matters
This development represents a substantial leap in AI code generation speed and efficiency, which could significantly accelerate software development. OpenAI's strategic shift to Cerebras chips also challenges Nvidia's market dominance and diversifies the high-performance AI hardware landscape.
Key Intelligence
- ■OpenAI has introduced GPT-5.3-Codex-Spark, a new model for real-time code generation.
- ■This new model is 15x faster than its predecessor, GPT-5.3-Codex, generating code at up to 1,000 tokens per second.
- ■Significantly, GPT-5.3-Codex-Spark is the first OpenAI model to deploy Cerebras chips, marking a major strategic move beyond Nvidia's hardware.
- ■The model offers 128k context and is currently available as a research preview for ChatGPT Pro users.
Source Coverage
Google News - Foundation Models
2/12/2026OpenAI's new Spark model codes 15x faster than GPT-5.3-Codex - but there's a catch - ZDNET
OpenAI Blog
2/12/2026Introducing GPT-5.3-Codex-Spark
Google News - AI & VentureBeat
2/12/2026OpenAI deploys Cerebras chips for 15x faster code generation in first major move beyond Nvidia - VentureBeat
Google News - AI & Bloomberg
2/12/2026OpenAI Debuts First Model Using Chips From Nvidia Rival Cerebras - Bloomberg.com
Google News - AI & Models
2/12/2026OpenAI Just Rolled Out Its First AI Model not Powered by Nvidia Chips - TipRanks
Google News - AI & TechCrunch
2/12/2026A new version of OpenAI’s Codex is powered by a new dedicated chip - TechCrunch
Google News - Foundation Models
2/12/2026Introducing GPT‑5.3‑Codex‑Spark - OpenAI
Google News - AI & Models
2/12/2026