GPT-5.3
Shares rank 32 of 72 language models by recommenders
People recommend it for fast, strong coding help and backend work, and for handling complex tasks well. Critics say the Instant variant is verbose, slow, and sometimes hallucinates.
People recommend it for fast, strong coding help and backend work, and for handling complex tasks well. Critics say the Instant variant is verbose, slow, and sometimes hallucinates.
58 people gave an opinion
- recommend it
- 26
- criticize it
- 32
- approval share
- 45%
Recommended in 8 of 13 months
- November 2025: not recommended
- December 2025: not recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: not recommended
- September 2026: not recommended
- October 2026: recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 26 people recommend itNeeds 10 or more, met.
- 0.8 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 8 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Strong for coding and agents
“Since Opus 4.6 and GPT 5.3, I have been able to absolutely crush my coding work”
Comment from 13 Jun 2026 in “I Think They [Anthropic] Are Lying to You [video]” “the coding inflection point was relatively recently (GPT 5.3 perhaps)”
Comment from 22 Jun 2026 on Hacker News “5.2 to 5.3 is the big leap for coding agents”
Comment from 6 Mar 2026 in “GPT-5.4 Thinking System Card” - Comment from 6 Mar 2026 in “GPT-5.4 Thinking System Card”
- Comment from 13 Feb 2026 in “I asked Claude Code to remove jQuery. It failed miserably”
Impressive on hard tasks
“that was a SOTA model only weeks ago. It blew people's minds about how good it was”
Comment from 25 Apr 2026 in “Changes to GitHub Copilot Individual Plans” “Both Opus 4.6 and GPT-5.3 one shot a Gameboy emulator for me.”
Comment from 5 Feb 2026 in “Claude Opus 4.6” “The huge jump from from Opus to GPT 5.3. Game changer.”
Comment from 4 May 2026 in “Redis array: short story of a long development process” - Comment from 6 Jul 2026 in “A global workspace in language models”
Fast responses
“GPT-5.3 Chat is the fastest model we've tested and it's still reasonably smart. Not meant for agentic tool usage / long context, though.”
Comment from 24 Apr 2026 in “DeepSeek v4” “when 5.3 is quite a bit faster”
Comment from 17 Mar 2026 in “GPT‑5.4 Mini and Nano”
What people warn about
Each mark is one comment
Poor code and hallucinations
“when we tried tuning Instant's persona directly it just chased warmth while still hallucinating”
Comment from 4 Mar 2026 in “GPT‑5.3 Instant” “GPT 5.3 tried to satisfy a typechecker by writing a BFS across an untyped object instead of just updating the type definitions.”
Comment from 31 Mar 2026 on Hacker News “Just yesterday I had gpt 5.3 generate completely awful code for the Cinema 4D Python API.”
Comment from 26 Jul 2026 in “What is happening to jobs? Separating AI hype from reality” Verbose, indirect, padded answers
“It still rambles, but now it prefaces it with "here's the short , to the point , direct answer:"”
Comment from 4 Mar 2026 in “GPT‑5.3 Instant” “I would be more impressed if it could give a straight yes/no answer to a yes/no question.”
Comment from 3 Mar 2026 in “GPT‑5.3 Instant” “Non-thinking models of OpenAI are insufferable yappers that don't know when to shut up”
Comment from 7 Mar 2026 in “GPT-5.4 Thinking System Card” - Comment from 4 Mar 2026 in “GPT‑5.3 Instant”
Instant variant weak and slow
“I can't get anything but 4 out of GPT 5.3 Instant, which should be the weakest available in the current generation.”
Comment from 3 May 2026 in “How many e's are in the word seventeen [video] (AI hallucination)” “wasn't even close to instant. Even with lowest effort it's like 3-4x best case TTFT compared to GPT-4.1”
Comment from 5 May 2026 in “GPT‑5.5 Instant”
Other language models people recommend
The four most recommended other language models, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Claude Opus | 3,331 | 2,746 | 55% | 13 | 4 mo156 owners | — |
| OpenAI GPT | 1,205 | 987 | 55% | 13 | 6 mo47 owners | — |
| Claude Sonnet | 854 | 586 | 59% | 13 | 6 mo46 owners | — |
| DeepSeek V4 Flash | 828 | 223 | 79% | 8 | 1 mo24 owners | Approved |
| GPT-5.3This page | 26 | 32 | 45% | 8 | — | — |
What people said
40 of the 59 comments that mention it, newest first. Each links to the original comment.
“it has been exceptionally well behaved with respect to following instructions”
“Just yesterday I had gpt 5.3 generate completely awful code for the Cinema 4D Python API.”
“the last model that was sufficiently competent and still open to discuss security”
“It just barely beat GPT-5.3 and was roughly equal with Sonnet 4.6.”
“they all found it instantly with search disabled”
“5.3 is remarkably useful still but you need to hold its hand”
“5.3 was incredibly better than 5.4/5.5”
“the coding inflection point was relatively recently (GPT 5.3 perhaps)”
“Since Opus 4.6 and GPT 5.3, I have been able to absolutely crush my coding work”
“wasn't even close to instant. Even with lowest effort it's like 3-4x best case TTFT compared to GPT-4.1”
“The huge jump from from Opus to GPT 5.3. Game changer.”
“I can't get anything but 4 out of GPT 5.3 Instant, which should be the weakest available in the current generation.”
“that was a SOTA model only weeks ago. It blew people's minds about how good it was”
“GPT-5.3 Chat is the fastest model we've tested and it's still reasonably smart. Not meant for agentic tool usage / long context, though.”
“I was trying to use GPT 5.3 for reverse engineering and got an account warning”
“I personally strictly prefer GPT 5.3+ models for backend work”
“Gemini 3.1 is better than GPT 5.3”
“GPT 5.3 tried to satisfy a typechecker by writing a BFS across an untyped object instead of just updating the type definitions.”
“when 5.3 is quite a bit faster”
“gpt-5.3 and 4 are very fine though.”
“Non-thinking models of OpenAI are insufferable yappers that don't know when to shut up”
“5.2 and 5.3 are strong/best for coding”
“5.2 to 5.3 is the big leap for coding agents”
“I just tried 5.3 and it gave me headings every 200 words. Absolutely infuriating.”
“when we tried tuning Instant's persona directly it just chased warmth while still hallucinating”
“It still rambles, but now it prefaces it with "here's the short , to the point , direct answer:"”
“The examples given in the video about "over-caveating" are more like bug fixes than enhancements to the model.”
“It also seems to do same or slightly worse than gpt-5.2-chat on my benchmarks.”
“I would be more impressed if it could give a straight yes/no answer to a yes/no question.”
“every time someone complains about it screwing up a task it turns out they were using the instant model”
“as apparently trying to describe "Auto but with Auto turned off" makes as little sense to them as it does to us.”
“GPT does much better than Claude in general”
“Instantly find confirmation bias for your illegal search & seizure of that ICE-protestor”
“works amazing on existing code bases in my experience”
“llms still really fail big in brand new challenges like i fail too”
“I ended up impressed enough w/ GPT 5.3 that I did the $200 for this month”
“GPT 5.2/5.3 and Opus 4.5/4.6 are better than Gemini 3 Pro.”
“The fact that they couldn't do it before is an indictment of how shit they were”
“Both Opus 4.6 and GPT-5.3 one shot a Gameboy emulator for me.”
“improve the quality using 5.3 in a couple months; best of both worlds”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.