GPT-5.4 Nano
OpenAI · released Mar 17, 2026
- Type
- Proprietary
- Input
- $0.20
- Output
- $1.25
- Cached
- $0.020
List price · per 1M tokens · OpenAI at 400K context · source ↗
Our take
Written Aug 3, 2026GPT-5.4 Nano is a small hosted model from OpenAI that can handle up to 400,000 tokens in a single request and accepts text, images and files. It is positioned as a budget option for long-context work, with coding as its standout measured strength.
Pick this for low-cost long-context tasks where you need to pass large documents or many files in one go. Use it for coding assistance, where it scores highest among its own capabilities, or for multimodal input on a budget. Skip it if you need strong creative writing output, if you want to self-host or avoid proprietary lock-in, or if throughput speed is critical and you cannot pay for the faster tier.
The case for it
- Extremely large request limit for its price tier: 400,000 tokens.
- Coding is the strongest measured capability, 58 points ahead of its own overall text score.
- Consistent scores across back-to-back evaluations, with under 0.6 points of drift across all six benchmark categories.
- The same provider offers a clear speed tradeoff: roughly double the cost for roughly double the throughput.
The case against it
- Creative writing is the weakest measured capability, 64 points below its own overall text score and 123 points below coding.
- Throughput at the cheapest tier is modest at 30.5 tokens per second, slower than both the provider's premium tier and some rival hosts.
- Proprietary weights with no disclosed licence, so self-hosting and redistribution are not possible.
How good is it?
IntelligencePuzzles, maths, exam questions
Arena Text (overall)73rd of 143 · 1402.4via High
CodingWriting and fixing code on its own
Arena Coding63rd of 143 · 1460.8via High
AgenticPlanning, calling tools, staying on task
GPT-5.4 Nano is not on Arena Agent (IPS), which is where the rating would come from, so there is no rating here. It is on LiveBench Agentic Coding, in 21st of 35 with 46.8.
WritingWe do not rate this
Two boards come close and neither tests writing: Arena Creative Writing asks people which of two replies they prefer, and LiveBench Language tests whether a model understood a passage. So we show where GPT-5.4 Nano placed and give it no mark out of five.
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, which is why they get no rating.
Every published score for this model14 scoresEvery figure we hold, from 14 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Or rent it from someone else
Why this differs from the header. The strip above quotes OpenAI's own list price. This is the cheapest live offer at the widest standard context we hold, whoever is serving it — a reseller undercutting a lab is ordinary commerce, not an error.
- per 1M tokens
- $0.20 in / $1.25 out
- Context served
- 400K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenAIflex tier | $0.10 / $0.63 | 400K | 63 tok/s | No | Yesunknown period | Unknown |
| OpenRouter | $0.20 / $1.25 | 400K | not measured | Unknown | Unknown | Unknown |
| Microsoft Azure AI | $0.20 / $1.25 | 400K | 35 tok/s | No | No | Confirmed |
| OpenAI | $0.20 / $1.25 | 400K128K out | 68 tok/s | No | Yesunknown period | Unknown |
Microsoft Azure AIAcross the 4 listings we hold: 3 say they do not train on prompts, 0 say they do and 1 do not say. 1 appear in the zero-retention registry we check; the rest are unknown to us rather than confirmed either way.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
- ✓
- Supported
- ✗
- Not supported
- Not published
- host gave no parameter list
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenAIflex | ✓ | ✓ | ✓ |
| OpenRouter | ✓ | ✓ | ✓ |
| Microsoft Azure AI | ✓ | ✓ | ✓ |
| OpenAI | ✓ | ✓ | ✓ |
Tool calling: 4 of 4 listings say yes. JSON output: 4 of 4 listings say yes. Strict schema: 4 of 4 listings say yes.
Models people weigh against GPT-5.4 Nano
When we formed this view
Dates behind this page
Prices last checked 9d ago
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 4 listings do not say whether they train on prompts.
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
Commercial API terms. We hold no licence record for this model, so there is nothing to summarise here.
Identifiers
- Modality record
- text+image+file->text
- Catalogue slug
- openai-gpt-5-4-nano