Gemini 2.5 Pro
Google · Released Jun 17, 2025
$1.25
Input price / MTok
#14 of 27$10
Output price / MTok
#20 of 271.0M
Context window
#11 of 28151
Tokens / sec
#8 of 2163.8
SWE-bench Verified
#11 of 1284
GPQA Diamond
#12 of 1986.7
AIME 2025
#6 of 1081.7
MMMU
#3 of 870.4
LiveCodeBench
#6 of 894.5
MRCR
#1 of 5Summary
Gemini 2.5 Pro is a proprietary language model from Google, released on Jun 17, 2025.
At $1.25 input and $10 output per million tokens, it ranks #14 of 27 for input price, placing it mid-range on input cost.
Its context window of 1.0M tokens ranks #11 of 28 for context depth in this dataset.
It targets long-context, reasoning, vision workloads.
Throughput is measured at 151 tokens per second (ranked #8 of 21 for throughput).
Now considered legacy as Gemini 3.x is GA.
Benchmarks
| Benchmark | Score | Max | % | Rank | Source |
|---|---|---|---|---|---|
| SWE-bench Verified | 63.8 | 100 | 63.8% | #11/12 | Google DeepMind blog (self-reported at l... |
| GPQA Diamond | 84 | 100 | 84.0% | #12/19 | Google DeepMind blog (self-reported at p... |
| AIME 2025 | 86.7 | 100 | 86.7% | #6/10 | Google DeepMind blog (self-reported at l... |
| MMMU | 81.7 | 100 | 81.7% | #3/8 | Google DeepMind blog (self-reported at l... |
| LiveCodeBench | 70.4 | 100 | 70.4% | #6/8 | Google DeepMind blog (self-reported at l... |
| MRCR | 94.5 | 100 | 94.5% | #1/5 | Google DeepMind blog (self-reported at l... |
Similar Models
Compare all 6 modelsModels with the most shared benchmark coverage and closest scores to Gemini 2.5 Pro.
Compare Gemini 2.5 Pro side by side
Pre-filled with 5 similar models
Speed source: https://artificialanalysis.ai/models/gemini-2-5-pro