GPT-5.3-Codex-Spark
Parameters
Undisclosed
Context Length
131K
Modality
Code
Architecture
Undisclosed
License
Proprietary
Release Date
12 Feb 2026
Knowledge Cutoff
-
API Pricing (per 1M)
-
Evaluation Benchmarks
No evaluation benchmarks for GPT-5.3-Codex-Spark available.
Rankings
Overall Rank
-
Coding Rank
-
About GPT-5.3-Codex-Spark
GPT-5.3-Codex-Spark is an ultra-low-latency coding model developed with Cerebras, optimized for live, interactive software engineering. Powered by the WSE-3 engine, it achieves over 1,000 tok/s generation throughput over a 128K context window.
Technical Specifications
Architecture specifications are undisclosed for proprietary models.
Resources
About GPT-5.3 Codex
GPT-5.3 Codex is OpenAI's ultra-fast coding model family, developed in partnership with Cerebras and the Wafer Scale Engine 3, purpose-built for real-time interactive coding with minimal latency.
Other GPT-5.3 Codex Models
- No related models available