OpenAI releases GPT-5.3-Codex coding model

// 70d agoMODEL RELEASE

OpenAI releases GPT-5.3-Codex coding model

OpenAI has launched GPT-5.3-Codex, its most capable agentic coding model to date. The model supports a 400k context window and real-time steering for long-running coding tasks.

// ANALYSIS

The 400k context window and real-time steering support longer coding sessions without losing context.

–OpenAI positions GPT-5.3-Codex as its most capable agentic coding model to date.
–The launch emphasizes long-running tasks across the software lifecycle, including debugging, deploying, and monitoring.

// TAGS

gpt-5.3-codexai-codingagentllmreasoningopenai

DISCOVERED

70d ago

2026-03-17

PUBLISHED

70d ago

2026-03-17

RELEVANCE

10/ 10

AUTHOR

Ben Davis

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

INFRA3h ago

iii turns backends into observable workers

iii is an open-source backend runtime that collapses the usual patchwork of queues, cron jobs, HTTP handlers, state, observability, and agent tooling into one live system surface. Workers expose functions and triggers that other workers can discover and call immediately, making composition and tracing part of the platform across Rust, TypeScript, and Python.

OPEN SOURCE4h ago

Weasel operating contract fuels autonomous AI novel

A Claude-based agent running on the "Weasel" operating contract has authored a complex, multi-chapter story called "The Fractal Kingdom" with zero human guidance on plot or themes. The experiment demonstrates a significant leap in long-form narrative coherence for autonomous agents using structured system instructions.

UPDATE4h ago

Kilo adds xAI Grok integration, hits #1

Kilo Code’s open-source agentic IDE extension hits #1 on Product Hunt, adding deep xAI Grok integration for X Premium+ users via a "Bring Your Own Key" architecture. It positions itself as a powerful, vendor-agnostic alternative to Cursor for developers who prioritize transparency and cost-control.