Is Hermes 4 (Thinking) or Gemma 4 31B Garnet cheaper?

Gemma 4 31B Garnet is cheaper on output tokens ($0.31 vs $0.40 per 1M).

Which has the larger context window, Hermes 4 (Thinking) or Gemma 4 31B Garnet?

Gemma 4 31B Garnet has the larger context window (262K tokens).

Hermes 4 (Thinking) vs Gemma 4 31B Garnet

Gemma 4 31B Garnet is cheaper on output tokens, while Gemma 4 31B Garnet offers a larger context window. Choose Hermes 4 (Thinking) or Gemma 4 31B Garnet based on the trade-off between cost, context, and the benchmarks that matter for your use case.

Spec	Hermes 4 (Thinking)	Gemma 4 31B Garnet
Provider	NanoGPT	NanoGPT
Input / 1M tokens	$0.20	$0.31
Output / 1M tokens	$0.40	$0.31
Context window	128K	262K
Parameters	—	—
Open weights	No	No
Released	Sep 2025	May 2026

Hermes 4 (Thinking) details →Gemma 4 31B Garnet details →

FAQ

Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.