Gemini 2.5 Pro Experimental 03-25

Name: Gemini 2.5 Pro Experimental 03-25
Author: Google Deep Mind

Reasoning modelGemini ProGemini 2.5

Release date: 2025-03-25Updated: 2025-04-21Views: 1,857

Live demoGitHubHugging FaceCompare

Parameters

No data

Context length

2000K

Multilingual

Supported

Reasoning ability

5/5

Gemini 2.5 Pro Experimental 03-25 is an AI model published by Google Deep Mind, released on 2025-03-25, for Reasoning model, and 2000K context length, with a 1393.40 score on Creative Writing.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Gemini 2.5 Pro Experimental 03-25

Model basics

Reasoning traces

Supported

Thinking modes

Thinking modes not supported

Context length

2000K tokens

Max output length

64K tokens

Model type

Reasoning model

Modality (in / out)

Text, Image, Audio, Video → Text

Release date

2025-03-25

Model file size

No data

MoE architecture

Total params / Active params

No data / Not applicable

Knowledge cutoff

No data

Gemini 2.5 Pro Experimental 03-25

Open source & experience

Code license

Proprietary

Weights license

Proprietary

GitHub repo

N/A

Hugging Face

N/A

Live demo

gemini.google.com

Gemini 2.5 Pro Experimental 03-25

Official resources

Paper

DataLearnerAI blog

N/A

Gemini 2.5 Pro Experimental 03-25

API details

API speed

4/5

No public API pricing yet.

Gemini 2.5 Pro Experimental 03-25

Benchmark Results

Gemini 2.5 Pro Experimental 03-25 currently shows benchmark results led by AIME 2024 (9 / 62, score 92), Aider-Polyglot (12 / 59, score 72.90), GeoBench ACW (5 / 20, score 81). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

General Knowledge

1 evaluations

Benchmark / mode

Score

Rank/total

HLE

Standard Mode

18.80

146 / 197

General Evaluation

1 evaluations

Benchmark / mode

Score

Rank/total

GPQA Diamond

Standard Mode

111 / 274

Common Sense

1 evaluations

Benchmark / mode

Score

Rank/total

SimpleQA

Standard Mode

52.90

13 / 47

Coding and Software Engineer

2 evaluations

Benchmark / mode

Score

Rank/total

LiveCodeBench

Standard Mode

70.40

57 / 128

SWE-bench Verified

Standard Mode

63.80

78 / 116

Math and Reasoning

3 evaluations

Benchmark / mode

Score

Rank/total

AIME 2024

Standard Mode

9 / 62

AIME2025

Standard Mode

86.90

46 / 107

FrontierMath - Tier 4

Standard Mode

4.20

40 / 80

Writing and Creative Capabilities

1 evaluations

Benchmark / mode

Score

Rank/total

Creative Writing

Standard Mode

1393.40

65 / 106

Common Sense Reasoning

1 evaluations

Benchmark / mode

Score

Rank/total

SimpleBench

Standard Mode

51.60

49 / 94

Agent Level Benchmark

2 evaluations

Benchmark / mode

Score

Rank/total

Aider-Polyglot

Standard Mode

72.90

12 / 59

BALROG

Standard ModeTools

43.30

6 / 12

Claw-style Agent Evaluation

2 evaluations

Benchmark / mode

Score

Rank/total

Claw Bench

Thinking ModeTools

80.40

20 / 29

Pinch Bench

Thinking ModeTools

71.90

30 / 38

Multimodal Understanding

2 evaluations

Benchmark / mode

Score

Rank/total

GeoBench ACW

Standard ModeToolsInternet

5 / 20

VPCT

Standard Mode

11 / 24

Long Context

1 evaluations

Benchmark / mode

Score

Rank/total

Fiction.liveBench

Standard Mode

66.70

13 / 16

View benchmark analysis Compare with other models

Gemini 2.5 Pro Experimental 03-25

Publisher

Google Deep Mind

View publisher details

Gemini 2.5 Pro Experimental 03-25

Model Overview

Gemini 2.5 Pro Experimental 03-25 is an AI model published by Google Deep Mind, released on 2025-03-25, for Reasoning model, and 2000K context length, with a 1393.40 score on Creative Writing.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.