Skip to content

AI model

GPT-5.3-Codex

OpenAI's GPT-5.3-Codex, a closed agentic coding model released in early 2026 for Codex and the API; its model ID is scheduled for API removal on April 1, 2027.

Known aliases

  • GPT-5.3
  • GPT Codex

Current stories

buildOne report1 publisher

Moving off three retiring OpenAI API models by April 2027 starts with finding every caller

OpenAI will remove gpt-5.3-codex, gpt-5.1 and gpt-5.4-nano from its API on April 1, 2027, according to a dev.to migration guide. The guide treats that date as the end of a staged move that starts by finding every caller, including the ones grep cannot see.

Publishers:dev.to

Reality

Evidence45
Adoption
Not scored
Hype gap0
Incentives
Not scored
Confidence50
buildConfirmed6 publishers

Anthropic's 26% figure grades Claude's R&D tasks by how much human steering they took

The lab says Claude leads 26% of its model research and collaborates on about 90%, with both tiers defined by how closely a human directs each task, and it wants rival labs publishing the same measure.

Perspective Coverage

6 publishers
Builder
Builder 39%
Operator
Operator 33%
Investor
Investor 28%

Reality

Evidence50
Adoption65
Hype gap+30
Incentives70
Confidence60
securityOne report1 publisher

OpenAI puts a GPT-5.4 reviewer where Codex used to stop and ask a human

Auto-review, shipped in Codex last week, hands escalation requests at the sandbox boundary to a separate GPT-5.4 Thinking call that approves about 99 percent of them and cuts human stops roughly 200-fold.

Publishers:alignment.openai.com

Reality

Evidence38
Adoption42
Hype gap+22
Incentives78
Confidence55
buildOne report1 publisher

A memory layer beat CLAUDE.md by 22.2 points, and 54 of 72 test pairs never moved

Deterministic checkers replaced the judge model, and the project instruction file finished below having no instruction file at all. The measurable part of context engineering turns out to be narrow.

Publishers:dev.to

Reality

Evidence44
Adoption10
Hype gap+26
Incentives78
Confidence40
investOne report1 publisher

GLM-5.3 Buys Buyers Time: Z.ai's Coding Model Cuts Tokens, Not the Closed-Model Lead

Z.ai says its 743B-parameter GLM-5.3 hits 34.5% on its own code bench using 22% fewer output tokens than GLM-5.2. The weights are still two weeks out.

Publishers:decrypt.co

Reality

Evidence32
Adoption18
Hype gap+38
Incentives76
Confidence36