GPT-5.4 Mini
OpenAI · released Mar 17, 2026
- Type
- Closed
- Input
- $0.75
- Output
- $4.50
- Cached
- $0.075
List price · per 1M tokens · OpenAI at 272K context · machine-readable source ↗
Our take
Written Sep 30, 2026GPT-5.4 Mini is a hosted-only model: we list no download for it, so using it means choosing a host. Its measured results are mid-pack on the Arena boards, with its strongest showing on set-piece coding and maths tasks and its weakest inside an agent harness.
Reach for it on set-piece code generation and completion, or on competition and olympiad-style maths, where its LiveBench figures are the strongest of its measured categories. It takes text, images and files, and its request capacity is large enough that a long report need not be split up first. Skip it if the job is agentic coding inside a harness, or if you meant to run the model on your own machine.
The case for it
- Competition and olympiad maths is its strongest measured category, at 78.46% on LiveBench with the xHigh variant.
- Code generation and completion follows at 71.62% on LiveBench with the xHigh variant, so set-piece coding work is where to start.
- A long report or a stack of documents fits beside the question, though reliable recall across all of it is unverified in our data.
- Images and files go into the same request as the text, so a screenshot or a document does not have to be described in words first.
The case against it
- Agentic coding inside a harness is its weakest measured area, at 41.67% on LiveBench with the xHigh variant.
- Mid-pack on the Arena boards: 52nd of 168 on Arena Text (overall) via High as of 25 Sep 2026, and 67th of 168 on Arena Creative Writing via High as of 25 Sep 2026.
- We list no download for it, so using it means choosing a host.
How good is it?
EverydayGeneral questions and everyday reasoning
Arena Text (overall)52nd of 168 · 1447
CodingWriting and fixing code on its own
Arena Coding57th of 168 · 1495
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent. It is on LiveBench Agentic Coding, in 51st of 58 with 41.67.
WritingDrafting and rewriting prose
Arena Creative Writing67th of 168 · 1402
Boards this model appears on that none of the ratings above are built on.
Every published score for this model15 scoresEvery figure we hold, from 15 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 between 59 min and 1 hour 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
- $0.75 in / $4.50 out
- Context served
- 400K
- Throughput
- ~46 tok/s
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenAIflex tierThrough OpenRouter | $0.38 / $2.25checked 1 hour ago | 400K128K max reply | 78 tok/s | No | Yesunknown period | Unknown |
| OpenRouterOpenRouter's own listing | $0.75 / $4.50checked 1 hour ago | 400K | not measured | Unknown | Unknown | Unknown |
| Microsoft Azure AIThrough OpenRouter | $0.75 / $4.50checked 1 hour ago | 400K128K max reply | 46 tok/s | No | No | Confirmed |
| OpenAIDirect | $0.75 / $4.50checked 59 min ago | 272K128K max reply | 66 tok/s | No | Yesunknown period | Unknown |
| Microsoft Azure AIusThrough OpenRouter | $0.82 / $4.95checked 1 hour ago | 400K128K max reply | 25 tok/s | No | No | Confirmed |
| OpenAIfast tierThrough OpenRouter | $1.50 / $9.00checked 1 hour ago | 400K128K max reply | 67 tok/s | No | Yesunknown period | Unknown |
OpenRouterAcross the 6 listings we hold: 5 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 AIusThrough OpenRouter | ✓ | ✓ | ✓ |
| OpenAIfastThrough OpenRouter | ✓ | ✓ | ✓ |
Tool calling: 6 of 6 listings say yes. JSON output: 6 of 6 listings say yes. Strict schema: 6 of 6 listings say yes.
Models people weigh against GPT-5.4 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 6 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-5-4-mini