GPT-5.4 nano
API ID gpt-5.4-nano
gpt-5.4-nano is an OpenAI model, released March 2026. It has a 400,000-token context window and up to 128,000 output tokens. It takes image input and supports function calling, structured outputs and reasoning. Standard price: from $0.20 input and from $1.25 output per 1M tokens.
- Status
- Deprecated
- Released
- March 2026
- Shutdown
- Apr 1, 2027
- Replacement
- GPT-6 Lunagpt-6-luna
It stops serving on April 1, 2027. Recommended replacement: gpt-6-luna.
Benchmarks
Scores as each source published themComparing with GPT-5 Nano, the previous model in this line. A score is compared only with the same benchmark from the same source, at the same setting where both have it. Full comparison: prices, limits, features
Arena
| Arena | GPT-5.4 nano | GPT-5 Nano | Change | Rank | Date | Source |
|---|---|---|---|---|---|---|
TextBy category
| 1401 ±4 | 1338 ±7 | +63 | 160 of 415 62,426 votes | Oct 13, 2026 | source |
| Text: Coding | 1460 ±6 | 1385 ±15 | +75 | 139 of 410 17,675 votes | Oct 13, 2026 | source |
| Vision | 1199 ±6 | 1146 ±11 | +53 | 93 of 165 25,650 votes | Oct 9, 2026 | source |
Rated as gpt-5.4-nano-high (high effort); GPT-5 Nano as gpt-5-nano-high (high effort). Ratings come from people voting between two anonymous answers; Text and Vision are style-controlled, as on arena.ai. Data: Arena (arena.ai), CC BY 4.0.
Epoch AI
| Benchmark | GPT-5.4 nano | GPT-5 Nano | Change | Setting | Run | Source |
|---|---|---|---|---|---|---|
| FrontierMath Tier 4 | 12.2% | not run | high effort | Jun 12, 2026 | source | |
FrontierMath Tiers 1-3Other settings
| 44.9% | not run | high effort | Jun 12, 2026 | source | |
GPQA DiamondOther settings
| 78.5% | not run | high effort | Apr 14, 2026 | source | |
OTIS Mock AIME 2024-2025Other settings
| 87.8% | not run | high effort | Apr 14, 2026 | source | |
| SimpleQA Verified | 11.7% | not run | high effort | Aug 27, 2026 | source |
Run independently by Epoch AI. Data: Epoch AI, Capabilities & benchmarking (epoch.ai/benchmarks), CC BY.
On cloud platforms
All platform dates| Platform | Platform ID | State | Shutdown | Time left | Replacement | Source |
|---|---|---|---|---|---|---|
| Microsoft Foundryversion 2026-03-17 | gpt-5.4-nano | Active | Sep 21, 2027 | 341 days | — | source |
Microsoft Foundry (version 2026-03-17) retires this model 173 days after OpenAI's own API.
Read daily from each platform's own pages. The shutdown date and countdown on this page use OpenAI's own date. Time left counts to 00:00 UTC on the shutdown date, which is the evening before in ET.
Alias
All aliasesCurrently points to gpt-5.4-nano-2026-03-17 (in place when tracking began (October 7, 2026)).
Features
Compare models| Feature | Value | Since | Source |
|---|---|---|---|
| Structured outputsThe model can be held to a JSON schema you supply, so every reply parses without retries or cleanup. | Supported | in place when tracking began (October 6, 2026) | source |
| Function callingAlso called tool use. You describe functions; the model replies with a function name and arguments for your code to run, then uses the result. | Supported | in place when tracking began (October 6, 2026) | source |
| Prompt cachingA prompt prefix you send again and again (system prompt, documents, tools) is cached, so later requests cost less and start faster. | Supported | in place when tracking began (October 6, 2026) | source |
| Batch APISend many requests as one job that runs in the background, usually finished within 24 hours, at a lower price than live requests. | Supported | in place when tracking began (October 7, 2026) | source |
| Image inputImages such as screenshots, photos and charts can go in the prompt for the model to read. | Supported | in place when tracking began (October 7, 2026) | source |
| Reasoning / thinkingThe model can work through a problem step by step before it answers: usually better answers to hard questions, at the cost of more tokens and time. | Supported | in place when tracking began (October 7, 2026) | source |
| Fine-tuningYou can train your own version of the model on your examples through the company's API, then call it like any other model. | Not supported | in place when tracking began (October 7, 2026) | source |
| Built-in web searchThe company's API can search the web for the model during a request and cite its sources; you don't need a search service of your own. | Supported | in place when tracking began (October 7, 2026) | source |
| Context windowThe most the model can work with in one request, in tokens: your prompt, conversation history and files. A token is about three-quarters of an English word. | 400,000 tokens | in place when tracking began (October 7, 2026) | source |
| Max output tokensThe longest reply the model can write in one response, in tokens. A token is about three-quarters of an English word. | 128,000 tokens | in place when tracking began (October 7, 2026) | source |
Not published by OpenAI: PDF input, temperature parameter, top_p parameter, logprobs.
Price history
USD per 1M tokens, standard tierWe started recording this model's prices on Oct 1, 2026. The history grows with each daily check.
Show as a table
| From | Type | Price |
|---|---|---|
| Oct 1, 2026 | Input | $0.20 |
| Oct 1, 2026 | Output | $1.25 |
Prices
USD per 1M tokens, every type and band, by tier| Type | Band | Price $/1M | In effect |
|---|---|---|---|
| Standard | |||
| input | short context | $0.20 | |
| cached input | short context | $0.02 | |
| output | short context | $1.25 | |
| Batch | |||
| input | short context | $0.10 | |
| cached input | short context | $0.01 | |
| output | short context | $0.625 | |
| Flex | |||
| input | short context | $0.10 | |
| cached input | short context | $0.01 | |
| output | short context | $0.625 | |
Changelog
Changes we detected on OpenAI's pages- Deprecated
OpenAI deprecated gpt-5.4-nano: shuts down Apr 1, 2027
Breaking
Related news
- DeprecationsBreakingOpenAI deprecated gpt-5.4-nano: shuts down 2027-04-01 (move to gpt-6-luna)
Counting to 8:00 pm ET on March 31, 2027, which is 00:00 UTC on April 1, 2027.
Email alerts
Get an email before gpt-5.4-nano shuts down
We'll email you 30, 7 and 1 days before it shuts down, and the same morning if the shutdown date moves or the price changes.
No password: one email to confirm, and you can unsubscribe any time. Privacy
Last checked · Sources: OpenAI deprecations, OpenAI pricing, OpenAI models