DeepHermes-3 Mistral 24B (Preview) vs Gemma 4 31B Garnet
DeepHermes-3 Mistral 24B (Preview) is cheaper on output tokens, while Gemma 4 31B Garnet offers a larger context window. Choose DeepHermes-3 Mistral 24B (Preview) or Gemma 4 31B Garnet based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | DeepHermes-3 Mistral 24B (Preview) | Gemma 4 31B Garnet |
|---|---|---|
| Provider | NanoGPT | NanoGPT |
| Input / 1M tokens | $0.30 | $0.31 |
| Output / 1M tokens | $0.30 | $0.31 |
| Context window | 128K | 262K |
| Parameters | — | — |
| Open weights | No | No |
| Released | May 2025 | May 2026 |
FAQ
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.