# OpenAI Decisions API

25 Hacker News commenters recommended OpenAI Decisions API and 19 criticized it, November 2025 to November 2026. It was recommended in 2 of 13 months.

**Verdict:** Developers recommend it as one fast, cheap API for choosing between options, far quicker than LLM calls. Critics say its confidence scores are unreliable.

## Why people recommend it

- Much faster than LLM calls (4 comments): "The main advantage is the speed (~2-300ms per input) which I really need for this purpose." ([source](https://news.ycombinator.com/item?id=50039088)) "It's much faster and cheaper (an order of magnitude)." ([source](https://news.ycombinator.com/item?id=49989037)) "decisions is 10x faster than responses API" ([source](https://news.ycombinator.com/item?id=49986799))
- Cheap per token, no output fees (3 comments): "it's still very cheap, and if you have existing enterprise OpenAI relationship it's a lot easier to onboard something like this" ([source](https://news.ycombinator.com/item?id=49985235)) "With decisions models, you don't pay for output tokens." ([source](https://news.ycombinator.com/item?id=49986984)) "It's $0.10 per million tokens. Dirt cheap." ([source](https://news.ycombinator.com/item?id=49985042))
- One API for several models (2 comments): "I'd much rather integrate one API than deal with separate integrations, auth, billing, and quirks for every provider." ([source](https://news.ycombinator.com/item?id=49930381)) "Unified API simplifies model comparison and integration." ([source](https://news.ycombinator.com/item?id=49904952))

## What people warn about

- Confidence scores seem unreliable (3 comments): "The probabilities I am seeing so far do not correspond with figures the business would find very agreeable" ([source](https://news.ycombinator.com/item?id=49993906)) "Luna cosplaying a decisions model and pretending the confidence score isn’t just a hallucination" ([source](https://news.ycombinator.com/item?id=50026128)) "it tends to return either 0 or 1 and doesn't return middle confidence levels very much at all" ([source](https://news.ycombinator.com/item?id=50027019))
- Biases and order effects like LLMs (2 comments): "These models seem to have the same biases and limitations of LLMs minus speed." ([source](https://news.ycombinator.com/item?id=49992223)) "It's still a predictive model under the hood, so order will always matter." ([source](https://news.ycombinator.com/item?id=49993887))
- Not better than LLM calls (5 comments): "this new method is much, much nicer but doesn't open up brand new use cases from what I see" ([source](https://news.ycombinator.com/item?id=50000891)) "It seems a bit silly since OpenAI LLMs already can output structured data." ([source](https://news.ycombinator.com/item?id=49898545)) "it was worse in all dimensions (price, speed, and quality of the result)" ([source](https://news.ycombinator.com/item?id=50097239))
- Seen as marketing or rushed copy (4 comments): "It's another Jev copy, like we've seen so many over the last few weeks." ([source](https://news.ycombinator.com/item?id=49899285)) "labelling something that is a simplification of existing models as" ([source](https://news.ycombinator.com/item?id=49989048)) "All about marketing that's why it's hard to root for Jev's success" ([source](https://news.ycombinator.com/item?id=49989705))

## Cite as

Hacker Approved, "OpenAI Decisions API", https://hackerapproved.com/p/decisions-api/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).