AIbase
EN
Home
AI Models Finder
O3 Mini
O

O3 Mini

A smaller version of O3, which is expected to perform better than previous generations of models in providing enhanced multimodal capabilities, improving inference capabilities, and more efficiently utilizing resources, while still maintaining strong performance on core tasks.
Intelligence(Relatively Strong)
Speed(Medium)
Input Supported Modalities
No
Is Reasoning Model
200,000
Context Window
100,000
Maximum Output Tokens
2023-09-30
Knowledge Cutoff
Go Compare

Pricing

￥7.92 /M tokens
Input
￥31.68 /M tokens
Output
￥13.86 /M tokens
Blended Price

Quick Simple Comparison

Input

Output

GPT-5‑pro
gpt‑oss‑120b
GPT-4.1 mini
￥0.4

Basic Parameters

o3-miniTechnical Parameters
Parameter Count
Not Announced
Context Length
200.00k tokens
Training Data Cutoff
2023-09-30
Open Source Category
Proprietary
Multimodal Support
Text Only
Throughput
115
Release Date
2025-01-31
Response Speed
141.0,872 tokens/s

Benchmark Scores

Below is the performance of o3-mini in various standard benchmark tests. These tests evaluate the model's capabilities in different tasks and domains.
Intelligence Index
65.51
Large Language Model Intelligence Level
Coding Index
56.61
Indicator of AI model performance on coding tasks
Math Index
92.23
Capability indicator in solving mathematical problems, mathematical reasoning, or performing math-related tasks
MMLU Pro
80.2
Massive Multitask Multimodal Understanding - Testing understanding of text, images, audio, and video
GPQA
77.3
Graduate Physics Questions Assessment - Testing advanced physics knowledge with diamond science-level questions
HLE
12.3
The model's comprehensive average score on the Hugging Face Open LLM Leaderboard
LiveCodeBench
73.4
Specific evaluation focused on assessing large language models' ability in real-world code writing and solving programming competition problems
SciCode
39.8
The model's capability in code generation for scientific computing or specific scientific domains
HumanEval
97.2
Score achieved by the AI model on the specific HumanEval benchmark test set
Math 500 Score
98.5
Score on the first 500 larger, more well-known mathematical benchmark tests
AIME Score
86
An indicator measuring an AI model's ability to solve high-difficulty mathematical competition problems (specifically AIME level)
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase