GPT-5.4

openai/gpt-5.4

Maker: OpenAI, United States
Released: 2026-03-05
Knowledge cutoff: 2025-08-31
Family: gpt
Weights: closed
Tasks: text.chat, text.reasoning, code, vision.understand, vision.extract
Cheapest offering: Azure OpenAI, about $0.007 for a typical text.chat call

Try it in the sandbox Use it from the API

Offerings

Provider	Model id at the provider	Status	Price per 1M tokens	Per-unit prices	Limits	Capabilities (level)
Azure OpenAI	`gpt-5.4`	gadeprecated 2027-09-02	Input: $2.5 Output: $15 Cache read: $0.25 Input, prompts over 272,000 tokens: $5 Output, prompts over 272,000 tokens: $22.5 Cache read, prompts over 272,000 tokens: $0.5 Input, priority tier: $5 Output, priority tier: $30 Cache read, priority tier: $0.5 Since 2026-11-03 · models.dev, MIT	none	Context: 1,050,000 tokens Max input: 922,000 tokens Max output: 128,000 tokens	`input.text` claimed `input.image` claimed `input.pdf` claimed `output.text` claimed `tools` claimed `structured_output` claimed `reasoning` claimed `parallel_tools` claimed `tool_choice` claimed `prompt_caching` claimed
Amazon Bedrock	`openai.gpt-5.4`	ga	Input: $2.75 Output: $16.5 Cache read: $0.275 Since 2026-11-03 · models.dev, MIT	none	Context: 1,000,000 tokens Max output: 128,000 tokens	`input.text` claimed `input.image` claimed `input.pdf` claimed `output.text` claimed `tools` claimed `structured_output` claimed `reasoning` claimed
OpenAI	`gpt-5.4`	ga	Input: $2.50 Output: $15 Cache read: $0.25 Input, prompts over 272,000 tokens: $5 Output, prompts over 272,000 tokens: $22.50 Cache read, prompts over 272,000 tokens: $0.50 Input, flex tier: $1.25 Output, flex tier: $7.50 Cache read, flex tier: $0.13 Input, priority tier: $5 Output, priority tier: $30 Cache read, priority tier: $0.50 Since 2026-11-03 · Curated by Nanner Tech	none	Context: 1,050,000 tokens Max input: 922,000 tokens Max output: 128,000 tokens	`input.text` claimed `input.image` claimed `input.pdf` claimed `output.text` claimed `tools` claimed `structured_output` claimed `reasoning` claimed `parallel_tools` claimed `tool_choice` claimed `prompt_caching` claimed
OpenRouter	`openai/gpt-5.4`	ga	Input: $2.5 Output: $15 Cache read: $0.25 Input, prompts over 272,000 tokens: $5 Output, prompts over 272,000 tokens: $22.5 Cache read, prompts over 272,000 tokens: $0.5 Since 2026-11-03 · models.dev, MIT	none	Context: 1,050,000 tokens Max input: 922,000 tokens Max output: 128,000 tokens	`input.text` claimed `input.image` claimed `input.pdf` claimed `output.text` claimed `tools` claimed `structured_output` claimed `reasoning` claimed

Quality scores

Basis	Score (0–1)	Raw	Measured	Source
`epoch:gpqa_diamond`	0.968	0.9330	2026-03-06	Epoch AI, CC-BY-4.0, normalized by Nanner Tech
`epoch:terminal_bench`	0.966	0.8180	2026-03-12	Epoch AI, CC-BY-4.0, normalized by Nanner Tech
`lmarena:vision_style_control`	0.939	1278.6393	2026-10-19	LMArena, CC-BY-4.0, normalized by Nanner Tech
`epoch:swe_bench_verified`	0.921	0.7686	2026-03-06	Epoch AI, CC-BY-4.0, normalized by Nanner Tech
`lmarena:document_style_control`	0.917	1470.0900	2026-09-13	LMArena, CC-BY-4.0, normalized by Nanner Tech
`epoch:frontiermath`	0.839	0.7860	2026-06-11	Epoch AI, CC-BY-4.0, normalized by Nanner Tech
`epoch:hle`	0.661	0.3624	2026-03-05	Epoch AI, CC-BY-4.0, normalized by Nanner Tech
`lmarena:webdev`	0.174	1401.0886	2026-10-28	LMArena, CC-BY-4.0, normalized by Nanner Tech

Aliases

Other names the API accepts for this model.

gpt-5.4
openai.gpt-5.4

Use it from the API

Look this model up, or ask for the best model for its task on the providers that offer it. See the quickstart and the API reference.

Look up this model

curl 'https://rhumbdata.com/v1/models/openai/gpt-5.4' \
  -H "Authorization: Bearer $RHUMB_KEY"

Resolve text.chat on its providers

curl 'https://rhumbdata.com/v1/resolve?task=text.chat&providers=azure-openai,bedrock,openai,openrouter' \
  -H "Authorization: Bearer $RHUMB_KEY"

Built from catalog version 2026-11-04.3 of 2026-11-04. Prices are list prices per 1M tokens unless marked otherwise; every score shows its source's notice (see Attributions).