Today we’re releasing the latest model in our reasoning series, OpenAI o3-mini, and you can start using it now in the API. o3-mini can outperform o1 in coding and other reasoning tasks, and is 93% cheaper and has lower latency. It supports function calling, Structured Outputs, streaming, and develop…

https://openai.com/careers/research-engineer/ << copy a job description Paste it into the input field: [Screenshot from 2025-02-01 21-56-31] and get courses from Microsoft Learn - connected to faiss… Build in ~1 hour. Obviously the search strategy is messed up a little

[image] edwinarbus: Would love to see! I think you should be able to put the URL here in the comments? Would love you to see… Here’s o3-mini giving its medium effort to fixing the “ using logprobs ” cookbook, and its long non-functional #4 output highlighter languishing. (Necessary as completi…

So, obviously we’ve an operational issue here. Multiple users already reported that in the playground they can access o3-mini, but via the API, they would get a 404. I can confirm the same thing as a tier 3 member. My implementation is correct – it works perfectly with o1-mini and definitely sets t…

It looks like o3-mini-high is really good at prompt-programs. In the code below, I’ve included a “memory-gram” prompt-program. When fed into o3-mini-high as a prompt, it adds a save feature. Saying “SAVE” automatically creates a memory-gram, which when fed into other instances of o3-mini-high reload…

Options for low or high in the API? How to set? (Loving the medium, BTW!) Answer - in case anyone else scrolls too fast: Field reasoning_effort in the request. Field usage.completion_tokens_details.reasoning_tokens in the response. Just make sure you also update your openai library! …

Where do I enable the o3 model in playgrounds? I can use o3 model just fine in chatgpt (Plus account) However I don’t see it in my dashboard in the playground or any ability to use it as a model in my API. Where do I enable it?

I don’t even see that model available in my playground? Any ideas why?

Tier 3 guy here, no access to it till date. Legitimate user, relying it on work deliverables. Had been applying multiple times since form opening in 2024. Spent hundreds of dollars on token so far. Got left out. Disappointing.

Same. Only can use it on ChatGPT unfortunately. They make formal announcement like it is launched for all, but truth is its opened only to a small subset of selected people. And again the selection criteria wasn’t revealed. OpenAI is really full of darkness. Hope o3-mini gets beaten instantly at all…

Launching o3-mini in the API

Announcements

Foxalabs February 2, 2025, 7:45am 54

please see Usage Tier 3 access denied to o3-mini - #42 by edwinarbus

Topic		Replies	Views
Usage Tier 3 access denied to o3-mini API	192	16055	April 17, 2025
Requesting o1-preview over the API API o1-preview	12	7054	December 10, 2024
New models and developer products announced at DevDay Announcements announcement	70	18557	February 16, 2024
AMA on the 17th of December with OpenAI's API Team: Post Your Questions Here Community api , ama , shipmas	193	8049	December 17, 2024
O3 is 80% cheaper and introducing o3-pro Announcements api , o3 , o3-pro	27	20227	July 31, 2025

Launching o3-mini in the API

Related topics