Gemini 3 Pro
Gemini 3 ProGoogleProprietaryNon-reasoningVerifiedReleased Nov 18, 2025
Overall score
61.2
Rank 53 of 222 · 50.4–72.0
Context window
2M
2,000,000 tokens
Price per 1M tokens
$2.00 in · $12.00 out
Source: BenchLM
Blended price
$4.50
3:1 input:output tokens
Output speed
—
Not measured yet
Category scores
Scores per capability area with the model's rank among ranked models in that category.
- 1Instruction following#4586.1
- 2Knowledge#4881.4
- 3Multimodal#1880.0
- 4Coding#6064.7
- 5Reasoning50.1
Benchmark results (26)
As reported in system cards, papers and public evaluations. Position compares this model with every other model reporting the same benchmark.
Multimodal
| MMMU-Pro | 81.0 | #9 of 48 |
| CharXiv | 81.4 | #22 of 42 |
| MMMU | 87.2 | #1 of 21 |
| MathVision | 86.6 | #12 of 20 |
| SimpleVQA | 73.2 | #5 of 17 |
| RealWorldQA | 83.3 | #9 of 16 |
| ERQA | 70.5 | #5 of 16 |
| ScreenSpot Pro | 72.7 | #10 of 15 |
| VideoMMMU | 87.6 | #2 of 12 |
| V* | 88.0 | #9 of 12 |
| OmniDocBench 1.5 | 88.5 | #8 of 11 |
| Video-MME (with subtitle) | 88.4 | #2 of 11 |
| MLVU (M-Avg) | 83.0 | #8 of 9 |
| CountBench | 97.3 | #2 of 8 |
| RefCOCO (avg) | 84.1 | #7 of 8 |
| Video-MME (w/o subtitle) | 87.7 | #1 of 7 |
| CC-OCR | 79.0 | #6 of 7 |
| AI2D_TEST | 94.1 | #1 of 7 |
| DynaMath | 85.1 | #4 of 6 |
| We-Math | 86.9 | #2 of 5 |
| MMLongBench-Doc | 60.5 | #3 of 5 |
| ODINW13 | 46.3 | #4 of 4 |
Agentic & tool use
| Gert Labs | 63.23 | #7 of 52 |
| JobBench | 11.4 | #31 of 34 |
Reasoning
| ARC-AGI-2 | 31.1 | #38 of 41 |
| ARC-AGI-1 | 75.00 | #33 of 36 |
Pricing by source
USD per 1M tokens.
| Source | Input | Output | Cached in |
|---|---|---|---|
| List price (BenchLM) | $2.00 | $12.00 | — |
Last published rate. Google's Gemini Developer API pricing page listed Gemini 3 Pro Preview at $2.00 input / $12.00 output per million tokens for prompts up to 200K tokens, rising to $4.00 / $18.00 above 200K. The pricing page, read September 25, 2026, no longer lists it.
Arena ratings
Human-preference ratings from LMArena. Best-rated variant shown.
| Text | 1479 | #20 · 42.1K votes |
| WebDev (coding) | 1440 | #78 · 14.1K votes |
| Vision | 1305 | #16 · 13.6K votes |
| Documents | 1434 | #33 · 10.8K votes |
Speed
Measured by us through OpenRouter (median of trials).
Not available on OpenRouter, so we cannot measure it yet.
Related models
Similar score
- Hy4 preview61.4 · $1.13
- Mistral Large 461.6 · $1.03
- GPT-5.3 Codex60.5 · $4.81
- GLM-5.260.5 · $2.15
- Gemini 3.5 Flash61.9 · $3.38