GPT-5.4
No shutdown before this date. GPT-5.4 stays on Amazon Bedrock until at least Jun 1, 2027.
≥ 8 mo- Earliest retirement
- ≥ 230 daysJun 1, 2027 · Amazon Bedrock
- Arena Text
- 1,475#21 of 128 · Arena
- GPQA Diamond
- 93.3%#15 of 108 · Epoch AI
- List price · $ per 1M
- $2.50 in $15.00 outOpenAI API · cached input $0.25
- Context window
- 1.1MMax output 128K tokens
Dates on every platform
3 listings · a floor is the earliest possible date
| Platform | Model ID and status | Deprecated | Shuts down | Replacement |
|---|---|---|---|---|
| OpenAI APIOpenAI: List models (GET /v1/models) (opens in a new tab) · checked Oct 14, 2026 | gpt-5.4Active · also gpt-5.4-2026-03-05 | – | None set | – |
| Azure AI FoundryMicrosoft: Model retirement schedule - Microsoft Foundry (opens in a new tab) · checked Oct 14, 2026 | gpt-5.4:2026-03-05Active | – | 11 mo | – |
| Amazon BedrockAmazon Web Services: Amazon Bedrock ListFoundationModels, us-east-1 (opens in a new tab) · checked Oct 14, 2026 | openai.gpt-5.4Active · also global.openai.gpt-5.4, us.openai.gpt-5.4 | – | ≥ 8 moNot sooner than | – |
In us-east-1, requests go through a cross-Region inference profile (listed in aliases).
What breaks when you move off GPT-5.4
No breaking changes are recorded for GPT-5.4.
Intelligence
Third-party results, credited below
| Benchmark | Publisher | Score | Rank | Published |
|---|---|---|---|---|
| Arena Text | Arena | 1,475 ± 2 | #21 of 128 | Oct 8, 2026 |
| GPQA Diamond | Epoch AI | 93.3% ± 1.8% | #15 of 108 | Mar 6, 2026 |
| Arena WebDev | Arena | 1,400 ± 6 | #44 of 69 | Oct 8, 2026 |
| Arena Agent | Arena | 0 ± 0 | #20 of 35 | Oct 12, 2026 |
| FrontierMath Tier 4 | Epoch AI | 49.0% ± 8.0% | #17 of 58 | Jun 11, 2026 |
| FrontierMath Tiers 1-3 | Epoch AI | 78.6% ± 2.4% | #17 of 71 | Jun 11, 2026 |
| OTIS Mock AIME 2024-2025 | Epoch AI | 97.8% ± 2.2% | #22 of 105 | Jul 15, 2026 |
| SimpleQA Verified | Epoch AI | 45.1% ± 1.6% | #33 of 68 | Aug 27, 2026 |
| SWE-bench Verified | Epoch AI | 76.9% ± 1.9% | #7 of 28 | Mar 6, 2026 |
Arena (arena.ai), https://arena.ai/leaderboard
Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource].
List prices
$ per 1M tokens, with each price’s source
| Platform | Input | Cached | Output | In effect | Source |
|---|---|---|---|---|---|
OpenAI APIgpt-5.4 | $2.50 | $0.25 | $15.00 | Oct 14, 2026Current · Requests over 272K input tokens are billed at higher long-context rates. Batch and Flex bill less; Fast mode bills more. | OpenAI: Pricing (opens in a new tab) |
Azure AI Foundrygpt-5.4:2026-03-05 | $2.50 | $0.25 | $15.00 | Mar 5, 2026Current · Global Standard rate from the US East 2 price list. | Microsoft: Azure Retail Prices API (Foundry Models, US East 2) (opens in a new tab) |
Amazon Bedrockopenai.gpt-5.4 | $2.75 | $0.275 | $16.50 | Oct 14, 2026Current · Standard tier, short context (272K input tokens or fewer). In-Region. Longer requests are billed at higher rates. | Amazon Web Services: Amazon Bedrock model card: GPT-5.4 (opens in a new tab) |