← Model list

GPT-5.4 Pro

openai·openai/gpt-5.4-pro·GA·Closed·gpt-pro series

More exact GPT-5.4 tier for demanding professional reasoning and agent tasks

At a glance
Official price$30.00 / $180 per 1M
Cache read · OpenAI
Lowest paid$27.00 / $160
Poe Gateway · 2.2× spread
Context1,050,000
⚠ Providers report 1,000,000–1,050,000; the table below is authoritative
Output limit128,000
Capabilities
ReasoningTool useStructured outputTemperatureAttachments
Modalities
TextImagePDF
Knowledge cutoff2025-08-31
Released / updated2026-03-05 / 2026-03-05

Quality & performance

Artificial Analysis doesn't cover this model (267 of 2059 have data). Quality data comes from independent evals covering widely used models.

Available at 18 providers18 with public prices

ProviderTierInputOutputCache readCache writeContextOutput limitStatus
PoeGateway$27.00$1601,050,000128,000
RequestyGateway$27.00$162$27.001,050,000128,000
ImpossiblGateway$30.00
tiered >272K: $60.00
$180$3.001,050,000128,000
OpenRouterGateway$30.00
tiered >272K: $60.00
$1801,050,000128,000
OpenAIOfficialFirst-party$30.00
tiered >272K: $60.00
$1801,050,000128,000
Cloudflare AI GatewayCloud$30.00
tiered >272K: $60.00
$1801,050,000128,000
OfoxGateway$30.00$1801,050,000128,000
Azure Cognitive ServicesFirst-party$30.00
tiered >272K: $60.00
$1801,050,000128,000
302.AIGateway$30.00
tiered >272K: $60.00
$180$0$01,050,000128,000
OpenCode ZenGateway$30.00$180$30.001,050,000128,000

Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.

Your usage cost

1302.AI$11,400.00
2Impossibl$11,760.00
3Eden AI$11,760.00
4Poe$13,400.00
5Requesty$13,500.00
6OpenRouter$15,000.00
7OpenAI · Official$15,000.00
Switch to 302.AI to save $3,600.00/mo (24%)
Note: this is a gateway; verify availability and rate limits yourself.

Benchmark11 items

NameConditionsScoreMetricSource
GPQA Diamond94.4accuracySource ↗
Humanity's Last Examvariant: no tools42.7accuracySource ↗
Humanity's Last Examvariant: with tools58.7accuracySource ↗
BrowseComp89.3accuracySource ↗
GDPval82wins or tiesSource ↗
FrontierMathdataset: Tier 1-350accuracySource ↗
FrontierMathdataset: Tier 438accuracySource ↗
ARC-AGI-1variant: Verified94.5accuracySource ↗
ARC-AGI-2variant: Verified83.3accuracySource ↗
FinanceAgentv1.161.5accuracySource ↗
GeneBench25.6accuracySource ↗

The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.

Reasoning control

effort = mediumeffort = higheffort = xhigheffort = noneeffort = loweffort = maxbudget_tokenseffort = minimal

Related models

Price historyone sample accumulated per data sync

Input listOutput listMin blended
$180$02026-08-052026-08-132026-08-05 · Input list $30.002026-08-06 · Input list $30.002026-08-07 · Input list $30.002026-08-08 · Input list $30.002026-08-09 · Input list $30.002026-08-10 · Input list $30.002026-08-11 · Input list $30.002026-08-12 · Input list $30.002026-08-13 · Input list $30.002026-08-05 · Output list $1802026-08-06 · Output list $1802026-08-07 · Output list $1802026-08-08 · Output list $1802026-08-09 · Output list $1802026-08-10 · Output list $1802026-08-11 · Output list $1802026-08-12 · Output list $1802026-08-13 · Output list $1802026-08-05 · Min blended $60.252026-08-06 · Min blended $60.252026-08-07 · Min blended $60.252026-08-08 · Min blended $60.252026-08-09 · Min blended $60.252026-08-10 · Min blended $60.252026-08-11 · Min blended $60.252026-08-12 · Min blended $60.252026-08-13 · Min blended $60.25