Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Google · Proprietary · rank 136 · bench-align-v5
Gemini 2.5 Pro is a general-purpose language model. It accepts text, images, PDFs, audio and video and produces text.
Gemini 2.5 Pro by Google scores 43.1 overall, costs $1.25 per 1M input tokens and $10.00 per 1M output tokens, and carries a context window of 1M tokens — observed 2026-11-21.
Seven axes from the ranking source. A missing axis is drawn as a gap.
Each axis shows the ranking source’s score and, beside it, BenchLM’s separate verified score for that axis. The two are published side by side and are not interchangeable.
Every published price component, including cache reads and writes.
The input, output and cache rates above are the base rate. models.dev publishes higher rates for requests whose prompt exceeds the thresholds below. The cost simulator prices a request whose whole prompt is above a threshold entirely at the highest tier it exceeds. The cost example on this page uses the base rate: its 1,200-token prompt is below every threshold.
Lifecycle events the source associates with this model, plus what this profile does not claim.
Measured values with the date they were observed. Nothing is inferred from a sibling model or a provider claim.
Time to first token: latency to the first answer chunk (Artificial Analysis, via BenchLM). Reasoning models include thinking time, so values can run to tens or hundreds of seconds.
Evidence key: Observed
| Measurement | Value | Observed | Last good | Evidence |
|---|---|---|---|---|
| Time to first token | 22.41 s | 2026-11-21 | 2026-11-21 | |
| Throughput | 118 tok/s | 2026-11-21 | 2026-11-21 |
Regional or per-endpoint measurements appear only when the API supplies them; none are modelled here.
10 conversations per day × 8 messages × 22 active days, 1200 input and 400 output tokens per message, no cache. This uses the same calculator as the cost simulator, so an unavailable applicable rate makes the example unavailable too.
Plans from the model's maker, and the API providers that list this exact model today. Links leave ModelScale; each provider sets its own prices and limits.
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| AA MMLU-Pro Artificial Analysis MMLU-Pro | 86.2 | #11 | Professional multi-subject questions | Professional knowledge and reasoning | Artificial Analysis MMLU-Pro Benchmark Leaderboard |
| GPQA Graduate-Level Google-Proof Q&A | 83 | #65 | 448 questions | Graduate level | GPQA: A Graduate-Level Google-Proof Q&A Benchmark |
| HLE Humanity's Last Exam | 18.8 | #69 | Expert-level questions | Frontier expert level | Humanity's Last Exam |
| Artificial Analysis Intelligence Index Artificial Analysis Intelligence Index | 16.1 | #130 | Cross-benchmark intelligence index | Display-only external reference | Artificial Analysis |
| AA-GPQA Diamond Artificial Analysis GPQA Diamond | 84.4 | #86 | Graduate-level science questions | Graduate-level science reasoning | Artificial Analysis GPQA Diamond Benchmark Leaderboard |
| AA-HLE Artificial Analysis Humanity's Last Exam | 22.5 | #110 | Expert-level questions | Frontier expert reasoning | Artificial Analysis Humanity's Last Exam Benchmark Leaderboard |
| AA-Omniscience Index Artificial Analysis Omniscience Index | -16.3 | #120 | Knowledge questions | Broad factual knowledge | AA-Omniscience: Knowledge and Hallucination Benchmark |
| AA-Omniscience Accuracy Artificial Analysis Omniscience Accuracy | 39.1 | #66 | Knowledge questions | Broad knowledge | Artificial Analysis model benchmarks |
| AA-Omniscience Hallucination Rate Artificial Analysis Omniscience Hallucination Rate | 90.9 | #30 | Knowledge questions | Factuality | Artificial Analysis model benchmarks |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| AA LiveCodeBench Artificial Analysis LiveCodeBench | 80.1 | #13 | Contamination-resistant coding tasks | Competitive programming | Artificial Analysis LiveCodeBench Benchmark Leaderboard |
| SWE-bench Verified Software Engineering Benchmark Verified | 63.8 | #84 | 500 verified issues | Professional software engineering | SWE-bench: Can Language Models Resolve Real-World GitHub Issues? |
| Vibe Code Bench Vibe Code Bench v1.1 | 0.40 |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| AA MATH-500 Artificial Analysis MATH-500 | 96.7 | #11 | 500 competition mathematics problems | High school to undergraduate mathematics | Artificial Analysis MATH-500 Benchmark Leaderboard |
| AIME 2024 American Invitational Mathematics Examination 2024 | 92 | #5 | 15 problems | High school olympiad level | American Invitational Mathematics Examination |
| FrontierMath v2 (Tiers 1-3) FrontierMath v2 Tiers 1-3 | 24.561 |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| AA-LCR Artificial Analysis Long Context Reasoning | 69.0 | #125 | Long-context reasoning tasks | Long-context reasoning | Artificial Analysis model benchmarks |
| CritPt Critical Physics Tasks | 2.6 | #106 | Research-level physics questions | Research-level physics reasoning | CritPt Benchmark Leaderboard |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| AA-IFBench Artificial Analysis IFBench | 48.7 | #85 | Verifiable instruction constraints | Instruction precision | Artificial Analysis IFBench Benchmark Leaderboard |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| AA Global-MMLU-Lite Artificial Analysis Global-MMLU-Lite | 90.3 | #9 | Multilingual knowledge questions | Multilingual professional knowledge | Artificial Analysis Global-MMLU-Lite Benchmark Leaderboard |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| GDPval-AA GDPval-AA | 616 | #107 | Agentic real-world work tasks | Professional agentic workflows | DeepSeek-V4 Technical Report |
| GDPval-AA GDPval-AA normalized | 0.0 | #105 | Economically valuable tasks | Professional agentic workflows | Artificial Analysis model benchmarks |
| AA Agentic Index Artificial Analysis Agentic Index | 3.5 |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| AA-MMMU-Pro Artificial Analysis MMMU-Pro | 74.9 | #58 | Multimodal academic reasoning | Frontier multimodal | Artificial Analysis MMMU-Pro Benchmark Leaderboard |
| Design Arena Website Design Arena Website Elo | 1156 | #84 | Website generation comparisons | Design and website generation | OpenRouter Grok 4.3 benchmarks |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| FailureSensorIQ FailureSensorIQ | 67.0 | #6 | Multiple-choice questions on sensors and failure modes for 10 industrial assets | Industrial maintenance and reliability knowledge | BenchLM |
| BinaryAudit BinaryAudit (backdoor detection in compiled binaries) | 21.1 | #11 | 19 backdoor and timebomb detection tasks on stripped binaries, three runs each (57 runs); 6 tooling and 8 clean-binary tasks reported alongside | Reverse engineering without source code or symbols | We hid backdoors in ~40 MB binaries and asked AI + Ghidra to find them (Quesma) |
| CompileBench CompileBench (building real projects from source, pass@1) |
A second, independent score source. Epoch’s own figures in Epoch’s units, one line per model version Epoch evaluated (effort setting as Epoch labels it). They are not part of this site’s capability scores or rank.
| Benchmark | Variant | Score | Epoch date | Evidence |
|---|---|---|---|---|
| GPQA Diamond Accuracy on graduate-level science multiple-choice questions (best score across scorers). Unit: fraction (0–1); higher is better. | default | 0.8529 ± 0.0212 | 2025-11-16 | |
| FrontierMath Tiers 1-3 (v2) Share of private research-level math problems solved, tiers 1-3. Unit: fraction (0–1); higher is better. | default | 0.2456 ± 0.0255 | 2026-06-11 | |
| FrontierMath Tier 4 (v2) Share of private research-level math problems solved, tier 4. Unit: fraction (0–1); higher is better. | default | 0 ± 0 | 2026-06-11 | |
| SWE-bench Verified Share of verified GitHub issues resolved (Epoch's own runs). Unit: fraction (0–1); higher is better. | default | 0.5756 ± 0.0225 | 2026-02-13 | |
| Terminal-Bench Terminal tasks completed; the model developer's own agent where Epoch lists one, otherwise the best listed agent (named beside the score), as compiled by Epoch from the public leaderboard. Unit: fraction (0–1); higher is better. | default Gemini CLI (developer's agent; 8 agents listed) | 0.196 ± 0.029 | 2025-11-04 |
Source: Epoch AI, CC BY 4.0 (licence), benchmark data read once a day.
People's votes between two anonymous answers on arena.ai, one board per task. LMArena's own ratings and ranks, one line per model version it rates; not part of this site's capability scores or rank.
| Board | Variant | Rating | Votes | Rank | Published | Evidence |
|---|---|---|---|---|---|---|
| Vision Arena People's votes between two answers about the same image, style-controlled. | — | 1,247 ± 5 | 88,464 | 59 | 2026-11-19 |
Source: LMArena, CC BY 4.0 (licence), leaderboard data read once a day.
| #43 |
| End-to-end web application builds |
| End-to-end software delivery |
| Vibe Code Bench: Evaluating AI Models on End-to-End Web Application Development |
| AA Coding Index Artificial Analysis Coding Index | 33.3 | #75 | Cross-benchmark coding index | Display-only external reference | Artificial Analysis model leaderboards |
| AA-SciCode Artificial Analysis SciCode | 46.3 | #71 | Scientific coding subproblems | Scientific programming | Artificial Analysis SciCode Benchmark Leaderboard |
| SWE-bench (Vals) SWE-bench, Vals AI run | 54.4 | #73 | Real repository issues by human time bucket | Frontier coding agents | Vals AI SWE-bench, Vals AI run leaderboard |
| #68 |
| 285 private advanced mathematics problems |
| From olympiad-plus to early research mathematics |
| FrontierMath v2 benchmark hub |
| FrontierMath v2 (Tier 4) FrontierMath v2 Tier 4 | 0.000 | #73 | 41 private extreme-difficulty mathematics problems | Research-level mathematics requiring hours or days of expert work | FrontierMath Tier 4 v2 leaderboard |
| #62 |
| Cross-benchmark agentic index |
| Display-only external reference |
| Artificial Analysis model leaderboards |
| Gert Labs Gert Labs Composite Game Benchmark | 42.01 | #34 | Novel game environments | Agentic coding and decision-making | Gert Labs rankings |
| τ²-bench results τ²-Bench Tool-Agent-User Evaluation | 54.1 | #95 | Airline, retail, and telecom customer-service task sets | Dual-control customer-service workflows | τ²-Bench: Evaluating Conversational Agents in a Dual-Control Environment |
| 53.3 |
| #11 |
| 15 build tasks on cowsay, jq, GNU coreutils and curl, three attempts each (45 attempts) |
| Dependency resolution, legacy toolchains, static linking and cross-compilation |
| CompileBench: can AI compile 22-year-old code? (Quesma) |
| GroundingME GroundingME: Multi-Dimensional Visual Grounding | 20.7 | #3 | 1,005 multi-dimensional visual grounding examples | Fine-grained object localization and ungroundable rejection | GroundingME: Exposing the Visual Grounding Gap in MLLMs through Multi-Dimensional Evaluation |