Inference
EN
Google

Gemini 2.5 Pro

Gemini 2.5 Pro is a standard model from Google. 20 benchmarks count toward its index, in 7 categories.

  • Closed weights
  • Standard
  • Input: text · image · file · audio · video
  • Released 2025-03
RankedEnough results for an index.
IndexAll categories, out of 100.46.6±4.1Median 51.0
CoverageIndex weight that has results.95%
SpeedOutput tokens per second.91/sMedian 67/s
First tokenSeconds to the first output token.2.9 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$1.25batch $0.625Median $0.60
Output / 1MUS dollars per 1M output tokens.$10.00batch $5.00 Batch price per 1M output tokens.Median $2.55
ContextMaximum tokens in one request.1.05MMedian 262K
EloLMArena rating and rank.1458±2 #42Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
google/gemini-2.5-pro
Max outputMaximum output tokens per request.
66K
Knowledge cutoffLast date of training data.
31 Jan 2025
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.125
WeightsLink to the published weights.
—
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
5 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
DatabricksDeprecatedCloud$1.25$10.001.05M
Google AI StudioMaker$1.25$10.001.05M
Google Vertex AIDeprecatedMaker$1.25$10.001.05M
OpenRouterDeprecatedRouter$1.25$10.001.05M
SAP AI CoreDeprecatedCloud$1.25$10.001.05M

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.5 counted40.4177 / 218
CodingCode writing and repair.5 counted48.8146 / 209
ReasoningLogic problems and puzzles.2 counted48.3129 / 227
MultimodalTasks with images and text.1 counted52.6—
KnowledgeFacts and expert knowledge.4 counted53.0112 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.1 counted34.5—
MathMath problems.2 counted46.0103 / 168

Results

20 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
American Invitational Mathematics Examination 2024Math92.0%5 / 27
Artificial Analysis Omniscience AccuracyKnowledge39.1%57 / 164
Terminal-Bench 1.0Agentic41.3% ±5.518 / 47
Artificial Analysis GPQA DiamondKnowledge84.4%82 / 199
Graduate-Level Google-Proof Q&AKnowledge83.0%60 / 141
SWE-bench Verified · swebench.comCoding53.6%20 / 46
IOI v1Coding17.1% ±6.227 / 61
Artificial Analysis Humanity's Last ExamKnowledge22.5%105 / 222
Critical Physics TasksReasoning2.6%102 / 214
Artificial Analysis MMMU-ProMultimodal74.9%56 / 113
Artificial Analysis Long Context ReasoningReasoning69.0%120 / 217
SWE-bench Verified · SWE-bench/experimentsCoding53.6%36 / 65
Artificial Analysis IFBenchInstruction48.7%81 / 146
Artificial Analysis SciCodeCoding46.3%68 / 114
Artificial Analysis MMLU-ProKnowledge86.2%9 / 15
τ²-Bench Tool-Agent-User EvaluationAgentic54.1%92 / 149
Gert Labs Composite Game BenchmarkAgentic42.0%34 / 52
Terminal-Bench 2.0 · Vals AIAgentic30.3% ±4.944 / 66
Software Engineering Benchmark VerifiedCoding63.8%81 / 114
GDPval-AA normalizedAgentic0.0%101 / 133
Humanity's Last ExamKnowledge18.8%64 / 77
τ²-bench BankingAgentic13.7%25 / 28
SWE-benchCoding54.4% ±2.281 / 87
Vibe Code Bench v1.1Coding0.4% ±0.4119 / 123
Artificial Analysis MATH-500Math96.7%9 / 9
Artificial Analysis LiveCodeBenchCoding80.1%12 / 12
MortgageTaxFinance68.9% ±0.98 / 97
MedCodeHealthcare50.6% ±2.117 / 106
CaseLaw v2Legal63.9% ±0.615 / 53
CorpFin v2Finance60.8% ±1.063 / 131
MedScribeHealthcare73.6% ±1.984 / 109
Finance Agent v1.1Finance41.6% ±2.840 / 51
Artificial Analysis Global-MMLU-LiteMultilingual90.3%—

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseVals AINo licence statedSWE-bench teamNo licence statedSierra ResearchMITmodels.devMIT

History

5 releases
Index46.7 → 46.6
23 Oct 202612 Nov 2026
Output / 1M$10.00 → $10.00
23 Oct 202612 Nov 2026

Questions

How much does Gemini 2.5 Pro cost?
$1.25 per 1M input tokens, $10.00 per 1M output tokens.
How good is Gemini 2.5 Pro?
Its index is 46.6 out of 100. That is place 172 of 283 ranked models.
What is Gemini 2.5 Pro best at?
Knowledge: 53.0 out of 100, place 112 of 268.
How fast is Gemini 2.5 Pro?
91 output tokens per second, 2.9 s to the first token.
What is the context window of Gemini 2.5 Pro?
1.05M tokens.
Is Gemini 2.5 Pro open source?
No. Its weights are not published.
When was Gemini 2.5 Pro released?
March 2025.

Badge

Gemini 2.5 Pro on Inference360: place 172 of 283, index 46.6
HTML
<a href="https://inference360.ai/models/gemini-2-5-pro"><img src="https://inference360.ai/badges/gemini-2-5-pro.svg" alt="Gemini 2.5 Pro on Inference360: place 172 of 283, index 46.6" height="20"></a>
Markdown
[![Gemini 2.5 Pro on Inference360: place 172 of 283, index 46.6](https://inference360.ai/badges/gemini-2-5-pro.svg)](https://inference360.ai/models/gemini-2-5-pro)

Compare with

Gemini 2.5 Flash (Preview)46.6Nemotron 3.5 Bolt 74B-A9B46.4Claude Sonnet 4 Thinking46.9Gemini 2.0 Flash29.5Gemma 3 27B25.3

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeNemotron 3 Ultra50.8 · Free

More from Google

All models
Gemini 2.0 Flash-LiteUnrankedGemma 3 27B25.3Gemini 2.0 Flash29.5Gemini 2.5 Pro Preview 06-05UnrankedGemini 2.5 Flash35.0