GPT-5.4

GPT-5.4 is an OpenAI flagship model for complex professional work, positioned as a more affordable option for coding and knowledge work.

start building →

compare models

                                   . .
                           ...::---+---::..
                       ..::+*=@@@@@@@@@%%=+:..
                     ..:+=@#@#####@#@#@@@#@@%+::. . ...... .
                   ..:+%@@@@@#@%%%====%@@@@#@@@*---++****+-:::.
                   .-=@###@%*+::..:..::-+*%@@@@@@@@@@@@@@@@@%=+-:.
                 ..+@@@#@=+:..         ..-+%@@@#@@@@@@@%@@@@@%@%%+:..
                 :+@@@@%+:..         .:-*%@@@@@@%%**++++***%%%@@@@%*:.
                :-@#@@%-.        ...:*=@@@@@@@=+-:..... ...:-+=%%%%%%+..
               .-%##@%-.       .:-+=@@@@@@%=+::.            ...-*%@@%%+:.
           .. ::*@#@@+..    ..:+=@@@@@@@=*::..                 ..+=%@%%*:
          .:-*=@@#@@=-.   .:+=@@@@@@@=*-..      ..:-+-:..       .:-=%%%%+.
       ..:*%@@@####@=:.   .+@@@@@@=*-:. .     .:-*%%%@%=-:.      ..-=%@%=:.
      .:*%###@#@@@#@*:    :*@@@@=-:...     ..-*%@@@@@@@%%%*-:.    ..*%%%%+..
     :+%@##@@@==@@@@=:.   :*@@@@+.     ...-*%@%@@@@%=%%%@@@%=*-..  .-=%%%*.
   .:+@###@%+-:+@@#@*:.  .:=@@@@-.  ...-*=%%@@@%%=+:::+*%%@@@@%=*::.:=%%%=:
  ..*@@#@@*-...-@@@@*:.  .:*@@@@-:.::+=@@@@@@@=*::.    .:+*%%%%%%%=*+=%%%=:
  :-%@#@@+:.  .-@@@@*:   ..*@#@@+-+=%@@@@@@@@%%+-...    ...:+*=%%%%%%%%%=*:
  :=@##@*:.   .-@@@@*:.  .:=@@@@%@%@@@@@==%%@%@%%=+:..       .:+*=%%%%%%%+.
 .+@@@#%-.    .-@@#@*:   .:*%@@@@@@@%=*::::+=%%%%%%%*-:...     ..:-==%%%%=-.
.:*@#@@+:     .-@@@@*:.   :*%@@@@%=+-:.    ..:+=%%@%%%=*-:.        .:+=%%%=+:
 :=@@#@-:     .-@@@@*:    :*@@@%*-:..       . .:-*%@@%%%%=*-:..     ..:*%=%=*:
 :=@#@@-:.    .-@@#@*:.  .:*%@@@-:              .-%%%%%@%%%%=+:.     ..:+=%%=*:.
.:*@@@@+:.    .-%@@@=-.. .:*@@@%:.              .:=%%%=*%%%%%%=*+:..    :+=%==+.
 .+@#@@=:.    .-%@#@@%*-:.:*@@@%-.              .:%%%%*-:+*=%%%%%=..    .:*%==*:.
 .:=@@@@+..    :*@@@@@@%=+-=%@%%-.              .:=%%%*: ..:+=%%%=:.    ..-=%=*-.
  :+@@@@%-.    .::+=@@@@@@%%@@@%:.              .:%%%=+:.   :+=%%=-.      :====+.
   .*@@@@%-..     .:-+=@@@@@@@@%-..             ::%%%%+.    .+=%%=:.      :====+.
    :+%@@@%*:.      ..::*=@@@@@@%*-:.        ..:+=%%%%*..   .-=%%=:.     .:====+..
     :+%@@@@%*-:...    ..:-*=@@%@%%=+-:.. ..:*==%%%%%=+.    .-=%==-.     .-===*-..
      .-*@#@@@@=*-:.       ..-*=%%%@@%=*--+=%%%%%%%%%=+:    .+====:.    .:*===*:
       :*@@@@@@@@@=*::.      ..:-*=%%%%%%%%%%%%=*+%%%%*..   .-====:.    .-*==*-.
      .:=%@@%%%@@@@@%=+:..      .:-*%%%@%%%%=*-:::%%%%+.    .-==%=:.  ..+====+.
      .:=@@@%+-+=%@@@@%%=+:. . .:+=%%%%%%=*+:.   :=%%%+.    .-===*-. .:+===*+:.
       :=%@@%-...:+=%%@@@%%=+-+*%%%@%%=*+:..    .:=%%=+..   .-===*-:-+*==**-..
      ..*@@@@+..  .:-*=%%%@@@%%%%%%%*+:..       .:====+..   .+====***====*-.
       .-%@@@=:.     .:-+=%%@%%%%=+:...     . .:+==%==+..   .-========**-:.
       .:=%@@%+:       ...-*===+-:..      .::+*=%=%%=*:     .-====***+:..
        .-%@%%%+:.       . .....       ..:+*===%===+-.     .:*==*+-:..
         .-%@%%%=-.                 ..:+==%%====*-:.      ..-*==*:. .
          .-=%@%%%*-.. .         ..:+*==%%%%=*-...        .:**=*-.
          ..:*%@%%%%=+-::......:-+*=%%%%%=*-.. .         .:**=*+:.
            ..:*=%@%%%%%%=****=====%%%=*+:...          .:-*===*:.
              .::+=%%%%%%%%%%%%%%%%%=%=-:..      .....:+*==**-:.
                  .:-+*==%=====***==%%%=**+-::...::-+**==***:..
                     .....:::.....:+==%=======*==**======*-..
                          .        .:-+==============*+-:..
                                     ...:--+**++++--:...
                                         . .......

input · usd / 1m tok

$2.50

output · usd / 1m tok

$15.00

context window

1.05mtokens

modalities · in → out

pricing

base api ratesusd per 1m tokens

input: $2.50over 272k · $5.00
output: $15.00over 272k · $22.50
cache read: $0.25over 272k · $0.50

brackets are by input tokens per request: a request over a threshold is billed at that bracket's rates for all of its tokens. cache creation can have a separate charge.

providers

Different companies host GPT-5.4. Impossibl routes your request intelligently and handles failover automatically when a provider is down.

Azure OpenAI
Azure AI Foundry
OpenAI

specifications

takes text and image in, gives text outopenai/gpt-5.4

creatorOpenAI
model typeText generation
context window1,050,000 tokens
maximum outputnot published
input
takes text and image in, gives text out
outputtext
tool callingnot verified
releasedmar 5, 2026
endpoint/v1/chat/completions

unknown means the public catalog does not establish that fact. a model's documented capabilities can differ from what a particular api supports. see the model documentation for integration details.

try it · your prompt, this model, live

# responses API
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["IMPOSSIBL_API_KEY"],
    base_url="https://api.impossibl.com/v1",
)

resp = client.responses.create(
    model="openai/gpt-5.4",
    input="▌",
)
print(resp.output_text)

edit the prompt, then run

# responses API
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["IMPOSSIBL_API_KEY"],
    base_url="https://api.impossibl.com/v1",
)

resp = client.responses.create(
    model="openai/gpt-5.4",
    input="▌",
)
print(resp.output_text)

edit the prompt, then run

benchmarks

intelligence index

39.0#20 of 90 models on the gateway

coding index71.1#22 of 60
omniscience5.8#25 of 86
gpqa92.0%#23 of 88
humanity’s last exam43.7%#13 of 89
terminal-bench v2.178.3%#28 of 57

GPQA92.0%
Tau287.1%
AA-LCR82.0%
MMMU-Pro78.4%
Terminal-Bench v2.178.3%
IFBench73.9%
Terminal-Bench Hard57.6%
Humanity’s Last Exam43.7%
Tau Banking39.6%
APEX-Agents33.3%
CritPt23.4%

Intelligence Index38.98
Omniscience Index5.78
GDPval Elo1306.71
coding index (API)71.1
intelligence index (API)39

view source ↗all source data and breakdowns (json) ↗

each configuration is listed separately; these evaluations describe the tested model configuration, and performance and costs can differ on impossibl. unmeasured scores are omitted and api values keep their source units. last checked 2026-09-10.

compare GPT-5.4 with other models

frontier labs and open weights, best-ranked firstusd / 1m tok · in / out

anthropic/claude-fable-5-1$10.00 / $50.00compare →
meta/muse-spark-1.3$1.25 / $4.25compare →
zai/glm-5.3$1.40 / $4.40compare →
xai/grok-4.6$2.00 / $6.00compare →

impossibl models --all →

frequently asked questions

What is GPT-5.4?

GPT-5.4 is an OpenAI flagship model for complex professional work, positioned as a more affordable option for coding and knowledge work. It unified the Codex and GPT lines into one system, supports text and image input with configurable reasoning effort, and anchors a family of mini, nano, and Pro tiers.

How much does GPT-5.4 cost?

GPT-5.4 costs $2.50 per million input tokens and $15.00 per million output tokens through the Impossibl API. Cache reads cost $0.25 per million tokens. These base rates apply through 272,000 total input tokens per request. Above 272,000 input tokens, the long-context rates in the pricing table apply to the whole request. All rates are in USD.

view the pricing table ↑

What is the context length of GPT-5.4?

GPT-5.4 has a 1,050,000-token context window through the Impossibl API.

What inputs and outputs does GPT-5.4 support?

GPT-5.4 accepts text and image as input. GPT-5.4 returns text.

How do I use GPT-5.4 through an API?

GPT-5.4 is available through the Impossibl API. Use openai/gpt-5.4 as the model identifier with /v1/chat/completions. The quickstart explains how to create an API key and send your first request.

follow the api quickstart →

When was GPT-5.4 released?

OpenAI released GPT-5.4 on March 5, 2026.

Creator release announcement ↗

this page describes GPT-5.4 as listed by the impossibl public api. prices and availability come from that catalog, refreshed here every five minutes. this is an impossibl offering, not a comparison of independent hosting-provider quotes. model descriptions and release dates, where available, use the creator sources linked under the answers above.

view the source catalog ↗model as markdown ↗report a correction →