GPT-5 Mini
OpenAI · released Aug 7, 2025
- Type
- Closed
- Input
- $0.25
- Output
- $2.00
- Cached
- $0.025
List price · per 1M tokens · OpenAI at 272K context · machine-readable source ↗
Our take
Written Sep 17, 2026GPT-5 Mini is a hosted-only model from OpenAI: we list no download for it, so using it means choosing a host. Its quality is measured by human preference votes and one coding score inside a specific harness, so treat it as a candidate to trial on work you can check yourself.
Use it for hosted chat, document and image work where you want a strong preference-rated model and can pick among the listed hosts. It takes text, images and files, and its request capacity is large enough that long documents need not be split up first. Skip it if you need to download and run the model yourself, if you need measured coding fixes in an existing codebase without review, or if you need disclosed licence terms before you build on it.
The case for it
- Rated on Arena leaderboards for coding, creative writing, hard prompts, instruction following, maths and overall text, so it is a reasonable first trial for mixed prompt work.
- Resolves 56.2% of real GitHub issues end-to-end on SWE-bench Verified, measured inside the mini-SWE-agent harness, so the figure is for the model in that harness rather than the model alone.
- Request capacity is large enough that long documents need not be split up first, though reliable recall across all of it is unverified in our data.
- Accepts text, images and files, so a screenshot or a document does not have to be described in words first.
The case against it
- We list no download for it, so using it means choosing a host rather than running it yourself.
- No licence is supplied, so what you may do with the model is unverified in our data.
- Arena scores are human pairwise votes rather than correctness measures, and the only coding score is harness-bound, so nothing here measures reasoning or maths correctness directly.
How good is it?
EverydayGeneral questions and everyday reasoning
Arena Text (overall)106th of 168 · 1390
CodingWriting and fixing code on its own
Arena Coding109th of 168 · 1431
Arena Coding is the only board that has scored it for this.
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent. It is on SWE-bench Verified, in 25th of 42 with 56.2.
WritingDrafting and rewriting prose
Arena Creative Writing116th of 168 · 1324
Arena Creative Writing is the only board that has scored it for this.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done.
Every published score for this model7 scoresEvery figure we hold, from 7 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Where to rent it
Prices checked 4 hours ago — each listing carries its own date.
Microsoft Azure AI, through OpenRouter
Why this differs from the header. The strip above quotes OpenAI's own list price; this is the cheapest live offer, whoever is serving it — a reseller undercutting a lab is ordinary commerce, not an error.
- per 1M tokens
- $0.25 in / $2.00 out
- Context served
- 400K
- Throughput
- ~98 tok/s
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenAIflex tierThrough OpenRouter | $0.13 / $1.00checked 4 hours ago | 400K128K max reply | 90 tok/s | No | Yesunknown period | Unknown |
| OpenRouterOpenRouter's own listing | $0.25 / $2.00checked 4 hours ago | 400K | not measured | Unknown | Unknown | Unknown |
| Microsoft Azure AIThrough OpenRouter | $0.25 / $2.00checked 4 hours ago | 400K360K max reply | 98 tok/s | No | No | Confirmed |
| OpenAIDirect | $0.25 / $2.00checked 4 hours ago | 272K128K max reply | 97 tok/s | No | Yesunknown period | Unknown |
| Microsoft Azure AIswedencentralThrough OpenRouter | $0.28 / $2.20checked 4 hours ago | 400K360K max reply | 50 tok/s | No | No | Confirmed |
OpenRouterAcross the 5 listings we hold: 4 say they do not train on prompts, 0 say they do and 1 does not say. 2 appear in the zero-retention registry we check; the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenAIflexThrough OpenRouter | ✓ | ✓ | ✓ |
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| Microsoft Azure AIThrough OpenRouter | ✓ | ✓ | ✓ |
| OpenAIDirect | ✓ | ✓ | ✓ |
| Microsoft Azure AIswedencentralThrough OpenRouter | ✓ | ✓ | ✓ |
Tool calling: 5 of 5 listings say yes. JSON output: 5 of 5 listings say yes. Strict schema: 5 of 5 listings say yes.
Models people weigh against GPT-5 Mini
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
What we do not know about this model yet
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 5 listings does not say whether it trains on prompts.
- We hold no batch or off-peak rate for any of its listings.
Licence and identifiers
What the licence allowsWe hold no licence record for this model. Inside are the identifiers you need to pull it — its Hugging Face repo where we have one, our slug and a machine-readable card.
Licence
We hold no licence record for this model, and no record of published weights either — so we can neither summarise its terms nor point you at the weights.
Identifiers
- Takes in, gives back
- Text, images and documents in, text out
- Catalogue slug
- openai-gpt-5-mini