Quoting Thibault Sottiaux
We’ve made GPT-5.3-Codex-Spark about 30% faster. It is now serving at over 1200 tokens per second.
— Thibault Sottiaux (https://twitter.com/thsottiaux/status/2024947946849186064), OpenAI
Author: npub1hvnsascw5fx...Published: 2/21/2026 01:45 AM Format: Markdown (kind 30023)
Identifier:
naddr1qvzqqqr4gupzpwe8pmpsagjvkjeu5y5um54zk5v0vl4622qnra8t4zqp86guwzl0qqvhzat0w35kueedw35xjcnpw4k8gttnda68g6tpw4uq498cts
We’ve made GPT-5.3-Codex-Spark about 30% faster. It is now serving at over 1200 tokens per second.
— Thibault Sottiaux (https://twitter.com/thsottiaux/status/2024947946849186064 ), OpenAI
Tags: openai (https://simonwillison.net/tags/openai), llms (https://simonwillison.net/tags/llms), ai (https://simonwillison.net/tags/ai), generative-ai (https://simonwillison.net/tags/generative-ai)
Comments (0)
No comments yet.