Gemma 4 31B Garnet vs Grok 4.3
Gemma 4 31B Garnet is cheaper on output tokens, while Grok 4.3 offers a larger context window. Choose Gemma 4 31B Garnet or Grok 4.3 based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Gemma 4 31B Garnet | Grok 4.3 |
|---|---|---|
| Provider | NanoGPT | SpaceXAI |
| Input / 1M tokens | $0.31 | $1.25 |
| Output / 1M tokens | $0.31 | $2.50 |
| Context window | 262K | 1M |
| Parameters | — | — |
| Open weights | No | — |
| Released | May 2026 | — |
| LiveBench score | — | 67.4 |
| LMArena (Chatbot Arena) Elo score | — | 1397 |
| LMArena WebDev Elo score | — | 1358 |
| LMArena Vision Elo score | — | 1229 |
| GPQA Diamond score | — | 88.8 |
FAQ
Is Gemma 4 31B Garnet or Grok 4.3 cheaper?
Gemma 4 31B Garnet is cheaper on output tokens ($0.31 vs $2.50 per 1M).
Which has the larger context window, Gemma 4 31B Garnet or Grok 4.3?
Grok 4.3 has the larger context window (1M tokens).
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.