GPT-5.4

openai/gpt-5.4

frontier model for general complex work.

cmd --model gpt-5.4
Intelligence index
51.4
Output speed
136.5 tok/s
Input
$2.50 /M
Output
$15 /M
Cache read
$0.25 /M
Agent-loop cost
$0.93 /M in
Context window
400K tokens
Released
March 5, 2026
Modalities
→

vs. the lineup

GPT-5.4 beside its stablemates and nearest rivals. The ◆ marks the best value in each column across every row shown.

pin a rival:
ModelIntelligenceCodingSpeedInput $/MOutput $/MBlended $/MContext
GPT-5.6 Sol58.9◆77.4◆64.8$5$30$11.251.05M◆
GPT-5.6 Terra5576.7122.3$2.50$15$5.631.05M◆
GPT-5.554.874.992.4$5$30$11.25200K
GPT-5.4 ◆51.471.1136.5$2.50$15$5.63400K
GPT-5.6 Luna51.271.4184.4◆$1$6$2.251.05M◆
GPT-5.3 Codex44.3—124.1$2$8$3.50400K
DeepSeek V4 Propinned44.359.470.9$0.43$0.87$0.541M
DeepSeek V4 Flashpinned40.356.2122$0.14◆$0.28◆$0.18◆1M
Compare all 50 models →

coding performance

The Intelligence Index and its sub-scores, ranked against every scored model in the catalog. A metric that has not been measured for GPT-5.4 has been left empty.

Coding Index
71.1
#12 of 34 scored
Terminal-Bench
78.3
#12 of 34 scored
Intelligence Index
51.4
#10 of 40 scored
Long-context reasoning
74
reasoning across a long context
SciCode
56.6
scientific coding
GPQA Diamond
92
graduate-level QA

usage calculator

How far a month of credits goes on GPT-5.4.

Input tokensfresh prompt
800
Output tokensmodel reply
180
Cache read tokensre-read context
50K
cost / request $0.017 · in $2.50 · out $15 · cache $0.25 per M
fresh input 12%output 16%cache reads 73%
Requests / 30 days
872
$15 credits ÷ $0.017 per request
~174 quick fixes~35 bug fixes~6 feature PRs

what real work costs

Real coding tasks priced end to end on GPT-5.4, from a quick lookup to a full-repo agent run.

One agent task
$0.33
180K in at 75% cache hit, 12K out
What you pay
$1.70 /M
all-in across every token that task touched
Sticker input
$2.50 /M
cache reads bill at $0.25 /M instead
TaskTokens in · outGPT-5.4Grok 4.5Claude Haiku 4.5
Quick lookup / one-liner8K · 1K$0.02$0.01$0.0065
Review a 500-line PR60K · 4K$0.12$0.08$0.04
Fix a bug (agent loop)180K · 12K$0.33$0.23$0.12
Refactor a module320K · 20K$0.58$0.41$0.21
Full-repo agent run900K · 45K$1.35$1.02$0.49

frequently asked

What is GPT-5.4 best for?+
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 text and image inputs, enabling high-context reasoning, coding, and multimodal analysis within the same workflow. The model delivers improved performance in coding, document understanding, tool use, and instruction following. It is designed as a strong default for both general-purpose tasks and software engineering, capable of generating production-quality code, synthesizing information across multiple sources, and executing complex multi-step workflows with fewer iterations and greater token efficiency.
How much does GPT-5.4 cost?+
$2.50/M input and $15/M output, cache reads $0.25/M. In an agent loop most input is cache-read, so the effective input rate is about $0.93/M.
Which plan do I need?+
Available on Pro and above.
How do I switch to it?+
Run cmd --model gpt-5.4, or type /model in a session and pick it. You can switch mid-session without losing context.
GPT-5.6 Sol→
frontier model for complex professional work · $5/$30 per M
GPT-5.6 Terra→
balances intelligence and cost · $2.50/$15 per M
All supported models→
sortable by score, speed and price

Ship code that matches your taste

Command Code is the AI coding agent that continuously learns your taste. Start for $1.

Start nowSee pricing
commandcode.ai/models/gpt-5-4