GPT-5.4

openai/gpt-5.4
VisionToolsJSONReasoning
by OpenAI · 2026-03-05

GPT-5.4 is OpenAI’s latest frontier model, unifying the Codex and GPT lines into a single system. It features a 1M+ token context window (922K input, 128K output) with support for...

ctx1.05M tokens
Max output128K
Inputtext + image + file
Outputtext
p50 TTFT6.60 s
INPUT$2.50/ 1M tokens
OUTPUT$15.00/ 1M tokens
p50 TTFT6.60 s7d
p95 TTFT10.00 s7d
TRAFFIC2.0Mtokens / 7d

GPT-5.4 is an OpenAI large language model with a context window of 1,050,000 tokens and max output of 128,000 tokens. It processes text, image, and file inputs. The model was evaluated on GPQA…

What is GPT-5.4?

Who should use GPT-5.4?

How does the context window affect performance?

What input modalities does it support?

Code samples

Call from any SDK

OpenAI-compatible — keep the SDK you already use

  • OpenAI SDKhttps://api.orcarouter.ai/v1
from openai import OpenAI

client = OpenAI(
    base_url="https://api.orcarouter.ai/v1",
    api_key="$ORCAROUTER_API_KEY",
)

response = client.chat.completions.create(
    model="openai/gpt-5.4",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Supported parameters

  • include_reasoning
  • max_completion_tokens
  • max_tokens
  • reasoning
  • response_format
  • seed
  • stream
  • structured_outputs
  • tool_choice
  • tools

Pricing

TierInput / 1M tokensOutput / 1M tokensCache read / 1M
272K$2.50$15.00$0.250
$5.00$22.50$0.500
Tier selected by input token count of each request

Cost calculator

Tokens / month10MM
Input share70%%
Estimated / month $62.50 · With prompt caching $54.63

Estimate based on list price

Tiered pricing — this estimate uses base-tier rates.

Token & cost estimator

Input tokens: 20Cost per request: $0.007550

Estimate only — actual token counts depend on the provider's tokenizer.

Performance

p50 TTFT
6.60 s
Output speed
71.7 tok/s
p95 TTFT
10.00 s
Error rate
0%

Public benchmarks

56.0
AA Coding
Better than 78% of models compared
#23 of 106
55.0
AA Intelligence
Better than 74% of models compared
#29 of 110
55.0
AA Math
Better than 31% of models compared
#56 of 81
GPQA Diamond
54.0 index
Humanity's Last Exam
41.6
IFBench
73.9
Long-Context Recall
74.0
MMLU-Pro
65.0 index
SciCode
56.6
TerminalBench Hard
57.6
τ²-Bench
47.0 index
Source: artificialanalysis.ai

How it compares

GPT-5.4GPT-5.2 ProGPT-5.4 ProGPT-5.5
Input $/M$2.50$21.00$60.00$5.00
Output $/M$15.00$168.00$270.00$30.00
Context1.1M400K1.1M
Quality9/1010/1010/1010/10
Compare side-by-sideCompare side-by-sideCompare side-by-sideCompare side-by-side

FAQ

What is the cost per token for GPT-5.4 on OrcaRouter?
Exact pricing is not provided in the facts. For current per‑token rates, please refer to OrcaRouter's official pricing page or contact their support. Costs scale with input and output token counts, so using the full 1,050,000‑token context will incur higher charges.
What is the context window of GPT-5.4?
GPT-5.4 has a context window of 1,050,000 tokens. This means the model can consider that many tokens of input (including text, images, and files) in a single request before generating a response up to 128,000 tokens long.
What are GPT-5.4's main strengths?
Based on provided facts, its main strengths are an extremely large context window (1,050,000 tokens), high score on GPQA Diamond (92.0), and support for multimodal inputs (text, image, file). This makes it ideal for tasks requiring deep reasoning over long documents or combining visual and textual information.
How does GPT-5.4 compare to GPT-4o?
GPT-5.4 has a vastly larger context window (1,050,000 vs. 128,000 tokens), higher maximum output (128,000 vs. 16,384 tokens), and a higher GPQA Diamond score (92.0 vs. not provided for GPT-4o). GPT-4o only supports text and image inputs, while GPT-5.4 also supports file uploads. GPT-5.4 is likely more expensive and slower.
How does OrcaRouter handle data when using GPT-5.4?
OrcaRouter does not train on customer data. Data handling practices follow industry standards. For specific compliance or privacy questions, review OrcaRouter's terms of service and data processing agreement. The API is accessed via HTTPS, and you can set data retention policies according to your needs.
How do I call GPT-5.4 via an OpenAI-compatible API on OrcaRouter?
Set your base URL to https://api.orcarouter.ai/v1 and use the model ID openai/gpt-5.4 in the request. Authenticate with your OrcaRouter API key. All standard chat‑completion parameters are supported. Refer to OrcaRouter's documentation for multimodal input formatting.
What input modalities does GPT-5.4 support?
GPT-5.4 accepts text, image (e.g., JPEG, PNG), and file (e.g., PDF, CSV) inputs. These can be combined in a single prompt. Images and files are processed as part of the context and count toward the token limit.
What is the maximum output length of GPT-5.4?
The model can generate up to 128,000 tokens in a single response. This allows for very long outputs, such as full book chapters or extensive code generation. Exceeding this limit requires multiple API calls or use of streaming.
Does GPT-5.4 support streaming?
Yes, the API supports a stream parameter (boolean). When set to true, responses are sent as server‑sent events. This is useful for real‑time applications. Streaming works with the standard chat‑completions endpoint.
How can I evaluate if GPT-5.4 is right for my use case?
Consider your context length needs: if you require more than 200,000 tokens or need to combine images and files, GPT-5.4 is a strong candidate. Test with representative prompts on OrcaRouter to measure accuracy, latency, and cost. For shorter or simpler tasks, cheaper models may suffice.

Embed this badge

OpenAI: GPT-5.4$2.50/M in6600ms p50via OrcaRouter
HTML <a href="https://www.orcarouter.ai/models/openai/gpt-5.4" target="_blank"> <img src="https://www.orcarouter.ai/embed/openai/gpt-5.4.svg" alt="OpenAI: GPT-5.4 on OrcaRouter" /> </a>
Markdown [![OpenAI: GPT-5.4](https://www.orcarouter.ai/embed/openai/gpt-5.4.svg)](https://www.orcarouter.ai/models/openai/gpt-5.4)

Model card as data

GET /api/public/models/openai/gpt-5.4Open
Machine-readable:/llms.txt/llms-full.txt