GPT-5.3-Codex

“GPT-5.5 Codex reasoning-token clustering at 516/1034/1552 may be leading to degraded performance on complex tasks”

“Some were definitely below par, and I recall having to iterate on the generated code more than I wanted to.”

“I have noticed this degradation of 5.5 reliability to what, in my experience, I consider Claude-level of reliability since early June.”

“occasionally it will seemingly short circuit and think for exactly 516 tokens, and return the wrong result”

“same prompting style, same repository documentation ala MD files, same skills, way different results”

“These stuck responses at fixed thresholds are strongly correlated with errors in complex tasks.”

“IMO somewhat weirdly 5.3-codex might be the best overall coding model OpenAI have ever released.”

“Somehow it's just way more careful than the others, and also much better at empirical verification of its hypothesis, writing tests, etc.”

“5.3 is good but talks like a robot, it’s too hard to understand what exactly it’s talking about.”

“subjectively think it does better job than Opus 4.6/4.7, with a fraction of the cost”

“its pricing is really good in production (due to how their token caching works)”

“GPT 5.5 codex was shockingly good. Achieved 90% perfect on first try in about a fourth of the time.”

“Gpt-5.3-codex is miles better than 5.4 in that regard. It's better at orchestration, and does the things that it said it did.”

“is really good about patiently watching and waiting on external processes like CI”

“5.3-codex was an improvement and 5.4 while _usually_ decent still regularly forgets, goes on unnecessary tangents”

“OpenAI's most recent GPT-5x codex model is better than Claude 5x for certain coding tasks”

“surprisingly adept at performing a high level feature branch review and catching issues”

“after GPT 5.3-Codex xhigh reached an impasse it went looking for it and found it!”

“when GPT-5.3 Codex looked at it, it said that speaker A and speaker B were switched in the LoRA code”

“it is providing working solutions on first go and following coding patterns that already exist in my app”

“I would personally be happy using gpt 5.3 codex for the foreseeable future, with just improvements in harnesses”

“Works well in both. My only complain is that it tends to get stutters after many sessions/hours.”

“gave me the confidence to load in a ridiculous number of tasks every day and let it rip”

“which has more than double context size, seems just as smart and costs the same”

“it's just so insanely good in almost everything else that I can look past these flaws”

“Since the performance of the latest models (Opus 4.6 max-effort, gpt-5.3-Codex) is nothing short of astonishing.”

“the Opus 4.5/4.6 and GPT-5.2/5.3 models really have represented a step-change in how good they are at running long tasks”

“I'm not seeing this translating to useful code. It failed with the first few requests I gave it.”

“getting their requests silently routed to a less capable model without telling the user at all”

“Pulling a switcheroo on the user behind the scenes, whatever the justification, is another issue”

“I have found GPT 5.3-Codex to do exceedingly well when working with graphics rendering pipelines.”

“When I've used codex (it has been getting better) but back in the day it would just do things and say it's done and you're just sitting there wondering "wtf are you doing?"”

“This is literally stuff of wonders and magic that redefines how we interface with computers and code.”

“Impressive jump for GPT-5.3-codex and crazy to see two top coding models come out on the same day”

“another model (preferably with a fresh session) can be a very effective peer reviewer”

“gpt codex 5.1 recently has been making a lot of typescript for me that is basically production ready”

“I've run multiple gpt-5 codex concurrent sessions in the cloud, but I didn't accept one thing they did.”

“GPT-5-Codex via OpenAI Codex CLi on Linux/WSL was, for me, the most capable model for coding”

“The latter is firmly preferred even by those who have no issue with budgeting tokens for their productivity.”

“The newer models are quite capable and in my experience can largely be treated like a co-worker for "most" problems.”