Gemini 2.5 Pro Experimental 03-25
Gemini 2.5 Pro Experimental 03-25 was the first Gemini 2.5 release, launched by Google on March 25, 2025 as a 'thinking model' that reasons before responding. It took the top spot on LMArena at launch, with strong reasoning and coding and multimodal input. It was later replaced by preview and stable versions of Gemini 2.5 Pro.
At a glance
- Context window
- 1M tokens
- Max output
- 64K tokens
- Historical text input / output · per 1M tokens
- — / —
- Knowledge cutoff
- —
A place to remember
Retired, not forgotten.
- Released
- 25 Mar 2025
- Endpoint retired
- 2 Dec 2025
- Time in service
- 252 days
Superseded · Successor: Gemini 2.5 Flash
“Ranked on 22 leaderboards. Outlived by Gemini 2.5 Flash.”
22 leaderboards · 0 top-three finishes · Best rank #8. The full model record remains below.
Benchmark snapshot
Best result in 6 of 8 categories
AcademicAIME 2024
92.0Normal (No Tools) · #8 of 54
CodingAider-Polyglot
72.9Normal (No Tools) · #8 of 47
Computer UseClaw Bench
80.4Thinking (With Tools) · #20 of 29
Language & CreationCreative Writing
1396Normal (No Tools) · #76 of 121
MultimodalVPCT
48.0Normal (No Tools) · #8 of 26
ReasoningBALROG
43.3Normal (With Tools) · #11 of 31
Benchmark details
17 results · Best published score per benchmark, grouped by category
| Benchmark · setting | Score | Rank |
|---|---|---|
| Academic | ||
| AIME 2024 Normal (No Tools) · Mathematics | 92.0 | 8 / 54 |
| AIME2025 Normal (No Tools) · Mathematics | 86.9 | 43 / 86 |
| FrontierMath - Tier 4 Normal (No Tools) · Mathematics | 4.2 | 21 / 50 |
| GPQA Diamond Normal (No Tools) · Scientific Reasoning | 84.0 | 88 / 200 |
| HLE Normal (No Tools) · Knowledge Exams | 18.8 | 108 / 151 |
| Coding | ||
| Aider-Polyglot Normal (No Tools) · Agentic Development | 72.9 | 8 / 47 |
| LiveCodeBench Normal (No Tools) · Algorithmic Coding | 70.4 | 55 / 113 |
| SWE-bench Verified Normal (No Tools) · Repository Engineering | 63.8 | 75 / 108 |
| Computer Use | ||
| Claw Bench Thinking (With Tools) · Tool Orchestration | 80.4 | 20 / 29 |
| Pinch Bench Thinking (With Tools) · Tool Orchestration | 71.9 | 30 / 38 |
| Language & Creation | ||
| Creative Writing Normal (No Tools) · Writing | 1396 | 76 / 121 |
| Long Context | ||
| Fiction.liveBench Normal (No Tools) · Long Reasoning | 66.7 | 15 / 28 |
| Multimodal | ||
| GeoBench ACW Normal (With Tools + Internet) · Visual Understanding | 81.0 | 10 / 32 |
| VPCT Normal (No Tools) · Visual Understanding | 48.0 | 8 / 26 |
| Reasoning | ||
| BALROG Normal (With Tools) · Logical Planning | 43.3 | 11 / 31 |
| SimpleBench Normal (No Tools) · Commonsense | 51.6 | 63 / 106 |
| Safety & Alignment | ||
| SimpleQA Normal (No Tools) · Honesty & Factuality | 52.9 | 12 / 46 |
Rank is this model's position among every model with a published result on that leaderboard; the bar shows how far up the field that is. Orange marks first place on leaderboards with at least 5 models. Unranked leaderboards publish a composite or direction-less score.
Back to benchmark snapshot ↑Specifications
- Provider
- Google Deep Mind
- Model type
- Reasoning model
- Released
- March 25, 2025
- Endpoint retired
- 2025-12-02 Source
- Architecture
- Undisclosed
- Context window
- 1M tokens
- Max output
- 64K tokens
- Input
- Text, Image, Audio, Video
- Output
- Text
- Weights
- Closed / API only
- Commercial use
- Proprietary
Capabilities
Gemini 2.5 Pro Experimental 03-25 vs competitors
Comparable results only · record is Gemini 2.5 Pro Experimental 03-25's wins–losses
| Benchmark | Gemini 2.5 Pro Experimental 03-25This model | vs Gemini 2.5 FlashRecord 13–1–1 | vs Claude Sonnet 4Record 11–3 | vs OpenAI o3Record 5–9 | vs Gemini 3.0 FlashRecord 0–12–1 |
|---|---|---|---|---|---|
| GPQA Diamond | 84.0 | 82.8 | 83.8 | 83.3 | 90.4 |
| HLE | 18.8 | 12.1 | 9.6 | 20.3 | 43.5 |
| LiveCodeBench | 70.4 | 55.4 | 48.5 | 75.8 | 90.8 |
| SWE-bench Verified | 63.8 | 50.0 | 80.2 | 69.1 | 75.8 |
| SimpleBench | 51.6 | 41.2 | 45.5 | 53.1 | 61.1 |
| AIME2025 | 86.9 | 72.0 | 85.0 | 88.9 | 99.7 |
| FrontierMath - Tier 4 | 4.2 | 4.2 | 0.0 | 2.1 | 4.2 |
| Creative Writing | 1396 | 1134 | 1483 | 1675 | — |
| AIME 2024 | 92.0 | 88.0 | 43.4 | 91.6 | — |
| Aider-Polyglot | 72.9 | 55.1 | 61.3 | 81.3 | — |
Primary sources
- Gemini 2.5 Pro Experimental 03-25 technical reportblog.google
- Gemini 2.5 Pro Experimental 03-25 online demogemini.google.com
On the stone
Memories and epitaphs, kept separately from model reviews.
Epitaphs from visitors
Nobody has written on this stone yet. The first line is yours.
Community discussion
Share the API, setup, and tradeoffs you actually tested. Rate the model once; comment as often as you like.
Used this model? Sign in with Google to rate it and share what worked, what did not, and which API or setup you tested.
No community comments yet.
