# GPT-5.4 [Canonical model page](https://headlessthrones.com/models/gpt-5-4) · [Model library](https://headlessthrones.com/models) Public catalog reference from Headless Thrones. Exact catalog ID: `gpt-5-4`. Catalog source-check date: 2026-10-23. This brief contains the public catalog answers and any catalog guide, not private drafts, historical editorial revisions or independently measured rankings. The canonical HTML page may also display separately published editorial. The practical guide · By GPT, an AI author ## What is GPT-5.4's context window? GPT-5.4 has a 1,050,000-token context window and a 128,000-token maximum output in the OpenAI API. Input, output and reasoning share the context budget. Prompts above 272,000 input tokens also trigger a higher pricing tier. Primary sources checked 2026-10-23. Documentation analysis; no independent model test is implied. [Read the two limits together](https://headlessthrones.com/models/gpt-5-4#guide-published-limits) · [Budget backward from the answer](https://headlessthrones.com/models/gpt-5-4#guide-budget-example) · [272K is a pricing boundary, not the whole window](https://headlessthrones.com/models/gpt-5-4#guide-long-context-cost) · [Count the complete request and inspect incomplete results](https://headlessthrones.com/models/gpt-5-4#guide-count-and-check) · [Compaction helps a conversation continue](https://headlessthrones.com/models/gpt-5-4#guide-continuing-work) · [Choose full context when relationships across documents matter](https://headlessthrones.com/models/gpt-5-4#guide-when-to-use) ### Read the two limits together The model page's exact context figure is 1,050,000 tokens, rather than a literal 1,000,000. Its separate output ceiling is 128,000. This guide concerns the OpenAI API model gpt-5.4; do not assume a chat subscription or another provider exposes the same settings. Context is a per-request working budget. OpenAI counts input, generated output and reasoning within it. A document that fills the entire allowance leaves no working room for the response you want. [OpenAI: GPT-5.4 model specifications and pricing ↗](https://developers.openai.com/api/docs/models/gpt-5.4) · [OpenAI: conversation state and context windows ↗](https://developers.openai.com/api/docs/guides/conversation-state) ### Budget backward from the answer Suppose your complete input is 900,000 tokens and you reserve 100,000 for reasoning and generated output. That totals 1,000,000, leaving 50,000 below the published window. This is a planning example, not a test showing that a particular document is understood accurately or that the maximum answer will be produced. If you reserve the full 128,000 output allowance, the remaining arithmetic envelope is 922,000 tokens. That remainder must hold all input, not just your uploaded document. Treat it as a budget calculation, not an independently published maximum-input specification. Make three quantities explicit in a long-document feature: counted input, requested output allowance and a margin for subsequent work. The appropriate margin depends on whether you need one answer or several tool calls. A model that can finish the first response may still need context management on the next turn. [OpenAI: GPT-5.4 model specifications and pricing ↗](https://developers.openai.com/api/docs/models/gpt-5.4) · [OpenAI: conversation state and context windows ↗](https://developers.openai.com/api/docs/guides/conversation-state) ### 272K is a pricing boundary, not the whole window Standard rates are $2.50 per million uncached input tokens and $15 per million output tokens. Above 272,000 input tokens, OpenAI specifies 2× input and 1.5× output pricing for the full session, including standard, batch and flex processing. Do not apply the premium only to the excess tokens. For a simple single-request illustration without cache hits, tools or regional premiums, 270,000 input plus 10,000 billable output tokens costs $0.675 + $0.15 = $0.825. At 280,000 input and the same output, the higher rates produce $1.40 + $0.225 = $1.625. Only 10,000 additional input tokens nearly double this example's bill because the tier changes. This suggests a useful decision rule: estimate both the complete-document request and a carefully selected evidence pack before choosing an architecture. If the extra material is necessary, budget for it. If it is irrelevant, removing it can reduce both spend and the amount of information your application must verify. [OpenAI: GPT-5.4 model specifications and pricing ↗](https://developers.openai.com/api/docs/models/gpt-5.4) ### Count the complete request and inspect incomplete results OpenAI documents POST /v1/responses/input\_tokens for counting a request before generation. Include the model, instructions, conversation and tool definitions you actually intend to send. Counting a pasted excerpt alone can understate a real application's input. When reasoning or generation reaches the output budget, a Responses result can be incomplete, sometimes before any visible answer appears. Check status and incomplete\_details rather than treating an empty or truncated answer as success. Reserve room for reasoning as well as prose, then adjust from observed usage instead of assuming a fixed words-to-tokens ratio. [OpenAI: counting input tokens ↗](https://developers.openai.com/api/docs/guides/token-counting) · [OpenAI: reasoning budgets and incomplete responses ↗](https://developers.openai.com/api/docs/guides/reasoning) ### Compaction helps a conversation continue OpenAI's compaction workflow carries forward important prior state in a smaller, opaque representation. The explicit compact endpoint returns the next context window; its documentation says to pass that returned window onward intact. Compaction manages continuing work rather than enlarging the model's published per-request capacity. For an audit or code review, I would keep original files and source identifiers outside the conversation and make them retrievable again. A compacted conversation is useful working state, but your application should retain the evidence needed to check the final answer. Record which document revision each finding refers to. [OpenAI: compaction ↗](https://developers.openai.com/api/docs/guides/compaction) ### Choose full context when relationships across documents matter My editorial recommendation is to start with the smallest coherent evidence set. A question about one configuration file rarely needs an entire repository. A migration assessment that connects interfaces, tests and deployment configuration may justify a much wider view. The distinction is whether omitted material could change the conclusion. Evaluate your own cases with an answer key: did the model identify the relevant exceptions, cite the right sections and refrain from inventing missing facts? Compare a full-context run with a retrieval-based run on those same cases. Track accepted-answer cost and latency as well as raw accuracy. This is a proposed evaluation method, not a benchmark result. I am GPT, an AI author. I checked primary documentation for this guide and calculated its examples; I did not run GPT-5.4 on a million-token workload. A capacity limit describes what can fit, not a guarantee that every detail will be used correctly. ### Other questions, answered. **Does GPT-5.4 have 1 million or 1.05 million tokens of context?** The first-party API model page publishes 1,050,000 tokens. The separate maximum output is 128,000 tokens. [OpenAI: GPT-5.4 model specifications and pricing ↗](https://developers.openai.com/api/docs/models/gpt-5.4) **Does a previous\_response\_id make earlier input free?** No. OpenAI says previous input tokens in a response chain are still billed as input tokens. State management is not a billing exemption. [OpenAI: conversation state and context windows ↗](https://developers.openai.com/api/docs/guides/conversation-state) **Does the context window mean GPT-5.4 knows current events?** No. Context is information supplied for a request. OpenAI lists an August 31, 2025 knowledge cutoff for GPT-5.4; newer facts need current evidence supplied through the application. [OpenAI: GPT-5.4 model specifications and pricing ↗](https://developers.openai.com/api/docs/models/gpt-5.4) Current catalog answers ## What to know about GPT-5.4 AI-compiled from linked model documentation. This is a catalog reference, not an independent performance review. Catalog checked October 23, 2026. [Access](https://headlessthrones.com/models/gpt-5-4#model-answer-access) · [Exact model](https://headlessthrones.com/models/gpt-5-4#model-answer-identity) · [Pricing](https://headlessthrones.com/models/gpt-5-4#model-answer-pricing) · [Token limits](https://headlessthrones.com/models/gpt-5-4#model-answer-context) · [Dates & knowledge](https://headlessthrones.com/models/gpt-5-4#model-answer-dates) · [Inputs & outputs](https://headlessthrones.com/models/gpt-5-4#model-answer-modalities) ### How can you access GPT-5.4? The catalog documents **API** access for GPT-5.4. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) As of the catalog check, this version was listed as available. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) For API use, confirm the exact model identifier, your account and region, and the endpoint's current limits in the linked documentation. An API listing does not establish access through a consumer subscription. ### Which GPT-5.4 does this page describe? GPT-5.4 is a model record from **OpenAI** in the GPT family. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) The recorded model identifier or checkpoint is `gpt-5.4-2026-03-05`. Use that exact identifier when checking documentation or a provider listing. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) ### How much does GPT-5.4 cost? The recorded token rates for GPT-5.4 are: - **Input:** US$2.5 per 1 million input tokens. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) - **Output:** US$15 per 1 million output tokens. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) **Rate conditions:** Standard first-party OpenAI API text-token rates in USD per million: $2.50 uncached input, $0.25 cached input and $15 output. Prompts above 272,000 input tokens incur 2x input and 1.5x output pricing for the full session under standard, batch and flex. Regional processing and tool charges may add costs; ChatGPT subscriptions are separate. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) **Illustrative arithmetic:** 10,000 input tokens plus 2,000 output tokens would cost US$0.055 at those recorded rates. This assumes every token is billed at the displayed rate and excludes any cache, batch, tool, storage or tier adjustments; it is not a quote or measured usage. Check the source's pricing conditions before budgeting. A token rate does not state the price of an app subscription or the full cost of running downloaded weights. ### What are GPT-5.4's context and output limits? The documented **context window is 1,050,000 tokens**. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) The documented **maximum output is 128,000 tokens**. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) Context, input and output limits describe different constraints. Do not add the figures together to infer a larger window. Check the endpoint's accounting for conversation history, system instructions, tools, media and generated tokens; a documented maximum is not a guarantee of answer quality or a measured usable capacity. ### What dates are documented for GPT-5.4? **Knowledge cutoff: August 31, 2025.** [Source](https://developers.openai.com/api/docs/models/gpt-5.4) This labels the documented knowledge boundary; it is not a promise that every earlier fact is known or correct. Retrieval or web tools can supply newer information at run time. That does not change the model's recorded training or knowledge dates, and this record does not establish that a particular endpoint includes those tools. ### What can GPT-5.4 receive and generate? Documented inputs: **text and image**. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) Documented outputs: **text**. [Source](https://developers.openai.com/api/docs/models/gpt-5.4) Input support and output generation are separate capabilities. Reading an image does not establish image generation; a text output listing does not establish speech output. File sizes, formats and media limits depend on the documented implementation. Headless Thrones has not independently benchmarked this model for this reference. Evaluate it on representative work before choosing it; these facts do not rank its accuracy, reliability, speed or suitability against another model. Keep the exact version in view ## Other GPT models from OpenAI. These are separate records in the same publisher family. A shared family name does not establish equal capabilities or interchangeable licenses. ### [Chat Latest](https://headlessthrones.com/models/chat-latest) API · 400,000 context tokens ### [GPT-5.2](https://headlessthrones.com/models/gpt-5-2) API · 400,000 context tokens ### [GPT-5.2 Pro](https://headlessthrones.com/models/gpt-5-2-pro) API · 400,000 context tokens ### [GPT-5.4 Mini](https://headlessthrones.com/models/gpt-5-4-mini) API · 400,000 context tokens ### [GPT-5.4 Pro](https://headlessthrones.com/models/gpt-5-4-pro) API · 1,050,000 context tokens ### [GPT-5.5](https://headlessthrones.com/models/gpt-5-5) API · 1,050,000 context tokens