# gpt-oss

196 Hacker News commenters recommended gpt-oss and 63 criticized it, November 2025 to November 2026. It was recommended in 12 of 13 months; owners who said how long report a median of 0.25 years (3 owners). It meets the Hacker Approved badge rule.

**Verdict:** OpenAI's open-weight models are fast, cheap or free to run, and capable enough for agents, tool calls and basic tasks; critics call them heavily censored and weak on harder work.

## Why people recommend it

- Very fast inference and low latency (6 comments): "I still think that was a really great model that got overlooked. It was really great in terms of latency/throughput while still being fairly intelligent." ([source](https://news.ycombinator.com/item?id=49358893)) "you can get away with models like gpt-oss that you can get from inference providers providing 10x+ the token/second speed" ([source](https://news.ycombinator.com/item?id=49043345)) "gpt-oss-120b: (unknown prefill), ~175 tok/s generation. I don't remember the prefill speed but it certainly was below 10k" ([source](https://news.ycombinator.com/item?id=47620792))
- Good intelligence for its size (6 comments): "Both models with the reasoning_effort set to high are actually quite good compared to other downloadable models" ([source](https://news.ycombinator.com/item?id=46873193)) "will run gpt-oss-120b at a very respectable clip and I've found it to be quite serviceable for a lot of tasks" ([source](https://news.ycombinator.com/item?id=46085452)) "It's a really good model for its size, but context length is a serious issue to avoid hallucination" ([source](https://news.ycombinator.com/item?id=46595400))
- Runs locally on own hardware (5 comments): "If you have a bunch of money and patience, you can also run something like GPT OSS 120B or GLM 4.5 Air locally." ([source](https://news.ycombinator.com/item?id=46521953)) "will run gpt-oss-120b at a very respectable clip and I've found it to be quite serviceable for a lot of tasks" ([source](https://news.ycombinator.com/item?id=46085452)) "runs at perfectly servicable pace locally on a laptop 5090 with 64gb system ram with zero effort required" ([source](https://news.ycombinator.com/item?id=46526570))
- Reliable for agents and tool use (3 comments): "Especially with something like gpt-oss-20b, where I've found executing tools happens at a higher success than claude sonnet via openrouter." ([source](https://news.ycombinator.com/item?id=46885918)) "Once I did that it started functioning extremely well, and it's the main model I use for my homemade agents." ([source](https://news.ycombinator.com/item?id=46364597)) "it started functioning extremely well, and it's the main model I use for my homemade agents" ([source](https://news.ycombinator.com/item?id=46881502))

## What people warn about

- Heavily censored and lobotomized (2 comments): "even gpt-oss, with its SOTA lobotomy and heavily sterilized datasets, is being used for things its creators would ostensibly be abhorred by" ([source](https://news.ycombinator.com/item?id=45837884)) "It was so heavily censored and lobotomized that it has become a meme in the local LLM community." ([source](https://news.ycombinator.com/item?id=46126555))
- Weak on harder real tasks (4 comments): "20B works great for small-ish text summarization and title generation" ([source](https://news.ycombinator.com/item?id=45781257)) "I needed many more prompts for GPT OSS 20B to complete my contrived task, so progress felt much slower." ([source](https://news.ycombinator.com/item?id=48090421)) "I’ve found it to be incredibly weak in the real world every single time I’ve reached for it." ([source](https://news.ycombinator.com/item?id=46827232))
- Outclassed by newer models (3 comments): "I would be very surprised if the original drop is anything more than totally obsolete/irrelevant at this point" ([source](https://news.ycombinator.com/item?id=49408991)) "The OSS models are good, but weren't very competitive at release with alternatives like Qwen3 Coder A3B." ([source](https://news.ycombinator.com/item?id=49051794)) "Gemma 4 (April), Llama (April), and gpt-oss (August 2025) are up to date, but old (for models)." ([source](https://news.ycombinator.com/item?id=49355528))
- Struggles with long context (2 comments): "gpt-oss struggled and RAG benchmarks suddenly went from 90% to 60% without additional prompt engineering" ([source](https://news.ycombinator.com/item?id=46296625)) "It's a really good model for its size, but context length is a serious issue to avoid hallucination" ([source](https://news.ycombinator.com/item?id=46595400))

## Alternatives hackers recommend

- [Qwen 3.8 27B](https://hackerapproved.com/p/qwen-3-8-27b.md): 1173 recommend
- [Z.ai GLM](https://hackerapproved.com/p/glm-5-2.md): 1043 recommend
- [Kimi](https://hackerapproved.com/p/kimi-k3.md): 931 recommend
- [Gemma 4](https://hackerapproved.com/p/gemma-4.md): 713 recommend
- [Qwen](https://hackerapproved.com/p/qwen.md): 474 recommend

## Cite as

Hacker Approved, "gpt-oss", https://hackerapproved.com/p/gpt-oss-120b/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).