The Open-Source AI Stack
RSS
All models

Models · o-series

OpenAI o4-mini

Proprietary OpenAI · 2025-04-16 · Proprietary

Released April 16, 2025 alongside o3 as a smaller multimodal reasoning model with a 200K-token context. Initially launched to paid subscribers and expanded to all ChatGPT users on April 24, 2025; OpenAI positioned it as the cost-efficient successor to o3-mini with image-input support.

Cost

$1.10 / Mtok input
$4.40 / Mtok output

OpenAI API · as of 2026-05-21

source ↗

Speed

151.2 tok/sec output
26323 ms TTFT

· as of 2026-05-21

source ↗

Architecture

tokens in Embedding vocab not disclosed × N layers Architecture not disclosed (proprietary or undocumented) Output projection tokens out
Schema-generated from data/models.yaml. Every label is auditable against the model's sources.

Specs

Architecture
unknown
Total params
not disclosed
Active params
not disclosed
Context window
200K tokens
Attention
unknown
Position encoding
unknown
Post-training
rlhf
OSI-approved
no
Data released
no
Training code
not released

Benchmarks

Each score carries the date it was published; we never infer or interpolate missing scores.

General reasoning

MMLU-Pro 83.2 as of 2026-05-21 source ↗
GPQA-Diamond 78.4 as of 2026-05-21 source ↗

Code

LiveCodeBench 85.9 as of 2026-05-21 source ↗

Math

MATH 98.9 as of 2026-05-21 source ↗
AIME 2024 94.0 as of 2026-05-21 source ↗
AIME 2025 90.7 as of 2026-05-21 source ↗

Available quantizations

None. The weights are not distributed, so there are no public quantizations.

Notable innovations

  • · Multimodal reasoning at o3-mini's price point
  • · Image-input chain-of-thought (analyzing diagrams and sketches)

Sources