GPT-5.4 Mini

OpenAI

balanced

Capabilities

VisionToolsCodeStructured OutputMultilingual

Strengths

Excellent cost/performance
Near GPT-5.4 on coding
Fast
400K context

Weaknesses

Not as strong on complex reasoning as full 5.4

Pricing

Input / 1M tokens

$0.75

Output / 1M tokens

$4.5

Context window

400k

Transparency

Open weights

0.0 / 10

Open training data

0.0 / 10

Open methodology

5.0 / 10

Licence openness

3.0 / 10

Provider disclosure

4.0 / 10

FMTI company score

350.0 / 10

Composite:2.4 / 10

OpenAI FMTI 35. Publishes system cards. No training data disclosure. Score dropped from 2024.

Sustainability

Inference energy

5.5 / 10

Training footprint

N/A

Provider infrastructure

4.0 / 10

Composite:4.8 / 10

Smaller model = more efficient. CEO stated 0.34 Wh per average ChatGPT query. Claims 100% renewable by 2027 via Microsoft Azure RECs (not direct PPAs). Zero published Scope 1/2/3 emissions as of 2026.

Task Fitness

question answering

79%

code

88%

mathematics

83%

business communication

68%

analysis

78%

extraction

85%

long-form generation

80%

research

55%

reasoning

78%

summarisation

82%

translation

80%

conversation

80%