Skip to content

Microsoft: Phi 4 Mini Instruct 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 Microsoft: Phi 4 Mini Instruct or Gemma 4 31B Garnet based on the trade-off between cost, context, and the benchmarks that matter for your use case.

SpecMicrosoft: Phi 4 Mini InstructGemma 4 31B Garnet
ProviderMicrosoftNanoGPT
Input / 1M tokens$0.08$0.31
Output / 1M tokens$0.35$0.31
Context window131K262K
Parameters
Open weightsNo
ReleasedMay 2026

FAQ

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