# GPT-5.3

26 Hacker News commenters recommended GPT-5.3 and 32 criticized it, November 2025 to November 2026. It was recommended in 8 of 13 months.

**Verdict:** 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.

## Why people recommend it

- Strong for coding and agents (5 comments): "Since Opus 4.6 and GPT 5.3, I have been able to absolutely crush my coding work" ([source](https://news.ycombinator.com/item?id=48510963)) "the coding inflection point was relatively recently (GPT 5.3 perhaps)" ([source](https://news.ycombinator.com/item?id=48629938)) "5.2 to 5.3 is the big leap for coding agents" ([source](https://news.ycombinator.com/item?id=47271316))
- Impressive on hard tasks (4 comments): "that was a SOTA model only weeks ago. It blew people's minds about how good it was" ([source](https://news.ycombinator.com/item?id=47897425)) "Both Opus 4.6 and GPT-5.3 one shot a Gameboy emulator for me." ([source](https://news.ycombinator.com/item?id=46904567)) "The huge jump from from Opus to GPT 5.3. Game changer." ([source](https://news.ycombinator.com/item?id=48012629))
- Fast responses (2 comments): "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." ([source](https://news.ycombinator.com/item?id=47893425)) "when 5.3 is quite a bit faster" ([source](https://news.ycombinator.com/item?id=47416774))

## What people warn about

- Poor code and hallucinations (3 comments): "when we tried tuning Instant's persona directly it just chased warmth while still hallucinating" ([source](https://news.ycombinator.com/item?id=47241673)) "GPT 5.3 tried to satisfy a typechecker by writing a BFS across an untyped object instead of just updating the type definitions." ([source](https://news.ycombinator.com/item?id=47593658)) "Just yesterday I had gpt 5.3 generate completely awful code for the Cinema 4D Python API." ([source](https://news.ycombinator.com/item?id=49056105))
- Verbose, indirect, padded answers (4 comments): "It still rambles, but now it prefaces it with "here's the short , to the point , direct answer:"" ([source](https://news.ycombinator.com/item?id=47241468)) "I would be more impressed if it could give a straight yes/no answer to a yes/no question." ([source](https://news.ycombinator.com/item?id=47240119)) "Non-thinking models of OpenAI are insufferable yappers that don't know when to shut up" ([source](https://news.ycombinator.com/item?id=47284490))
- Instant variant weak and slow (2 comments): "I can't get anything but 4 out of GPT 5.3 Instant, which should be the weakest available in the current generation." ([source](https://news.ycombinator.com/item?id=48001056)) "wasn't even close to instant. Even with lowest effort it's like 3-4x best case TTFT compared to GPT-4.1" ([source](https://news.ycombinator.com/item?id=48028580))

## Alternatives hackers recommend

- [Claude Opus](https://hackerapproved.com/p/claude-opus.md): 3331 recommend
- [OpenAI GPT](https://hackerapproved.com/p/gpt-5-5.md): 1205 recommend
- [Claude Sonnet](https://hackerapproved.com/p/claude-sonnet.md): 854 recommend
- [DeepSeek V4 Flash](https://hackerapproved.com/p/deepseek-v4-flash.md): 828 recommend
- [Fable](https://hackerapproved.com/p/fable.md): 823 recommend

## Where to get it

- [Visit the official site](https://openai.com/)

## Cite as

Hacker Approved, "GPT-5.3", https://hackerapproved.com/p/gpt-5-3/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).