Gemma 4 12B (Reasoning) is a Google Gemma open-model profile, aimed at developers who want deployable Google-developed models rather than hosted Gemini-only access. The benchmark data is especially useful for comparing its efficiency and capability against similarly sized open and open-weight alternatives.
Introducing GemmaRank #224 across 526
Rank #171 across 436
Rank #43 across 357
Percentile score by analysis domain.
* Cost is inverted: lower input, output, and blended prices rank higher.
Higher bars mean stronger relative placement.
| Metric | Domain | Value | Rank |
|---|---|---|---|
| Artificial Analysis Intelligence Index | overall | 22.0 | #224 |
| Artificial Analysis Coding Index | coding | 24.9 | #171 |
| GPQA | reasoning | 75.3% | #169 |
| Humanity's Last Exam | reasoning | 14.8% | #110 |
| SciCode |
| coding, reasoning |
| 38.2% |
| #149 |
| Output Speed | speed | 158.6 tok/s | #65 |
| Time to First Token | speed | 1.36s | #180 |
| Blended Price | cost | $0.150/M | #43 |
| Input Price | cost | $0.100/M | #44 |
| Output Price | cost | $0.300/M | #50 |
| Value Index | cost, overall | 146.7 | #34 |