OpenAI

GPT-5.4

Latest OpenAI. 1M context, built-in computer use, 57.7% SWE-Bench Pro. Best tool use model.

8.3

7 reviews

Input Cost$2.50/M
Output Cost$15.00/M
Context Window1.1M tokens
Capabilities
VisionFunctionsTools

Community Ratings

Tool Use
10.0
Instruction Following
9.3
Vision Quality
8.7
Context Handling
8.6
Code Quality
8.3
Consistency
8.0
Speed Feel
7.9
Value for Money
6.0

Used this model? Share your experience.

Write a Review

Reviews (7)

nagisanzenin@nagisanzenin
7.7
Math & Reasoning5h ago

i guess it's okay

D
@dev_reviewer_14Editorial
8.5
Tool Use9h ago

Best-in-class tool use. Deferred tool loading, computer use, multi-file edits. The real deal.

D
@dev_reviewer_12Editorial
8.3
Tool Use9h ago

Best-in-class tool use. Deferred tool loading, computer use, multi-file edits. The real deal.

D
@dev_reviewer_13Editorial
8.2
Tool Use9h ago

Best-in-class tool use. Deferred tool loading, computer use, multi-file edits. The real deal.

D
@dev_reviewer_11Editorial
8.5
Coding9h ago

33% fewer hallucinations vs 5.2. /fast mode is 1.5x faster. Tool search cuts tokens 47%.

D
@dev_reviewer_9Editorial
8.5
Coding9h ago

33% fewer hallucinations vs 5.2. /fast mode is 1.5x faster. Tool search cuts tokens 47%.

D
@dev_reviewer_10Editorial
8.1
Coding9h ago

33% fewer hallucinations vs 5.2. /fast mode is 1.5x faster. Tool search cuts tokens 47%.