Gemma 3 27B
Google released 2025-03-12
Gemma 4 31B is both better and cheaper.
It scores +22.3 higher and costs 7% less ($0.160/M against $0.172/M) on this workload — and it does everything this model does.
DeepSeek V4 Flash 0731 is cheaper still (39% less) but drops no image input.
Every price dimension
| Input | $0.080/M |
|---|---|
| Output | $0.450/M |
| Cached input | $0.040/M |
Independent scores
| Intelligence | 7.4 |
|---|---|
| Coding | 10.1 |
| Agentic | 0.3 |
| LMArena Elo | 1358.3default |
Sources are listed separately rather than averaged. Across the 62 models both have scored they correlate at r = 0.806 — close agreement overall, but four models rank very differently between them, and a blended score would hide exactly those.
Capability
| Context window | 262K |
|---|---|
| Max output | 131K |
| Input modes | text, image |
| Tool use | yes |
| Reasoning | no |
| Knowledge cutoff | 2024-08-31 |
Provenance
| Price source | openrouter.ai |
|---|---|
| Fetched | 2026-08-24 |
| Quality data | verified |
Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities,...