Gemma 4 31B API Prices Rise: Input Cost Jumps 122%

What Changed

Gemma 4 31B's API pricing has moved up sharply. Input tokens now cost $0.20 per million, up from $0.09 — a +122% increase. Output tokens rose more modestly, from $0.34 to $0.40 per million tokens.

The output side is a smaller jump in percentage terms, but the input price change is the one to watch. Most real-world workloads — especially RAG pipelines, document processing, or long-context tasks — are input-heavy, so that 122% increase will hit harder than the output number suggests.

Does This Matter for Your Workload?

At $0.09 input, Gemma 4 31B was positioned as a budget-friendly mid-size model. At $0.20, it's less of a standout deal. To put it concretely: if you were processing 100 million input tokens a month, your monthly input bill just went from $9 to $20.

That's still not expensive in absolute terms, but it does narrow the gap between Gemma 4 31B and pricier alternatives. If you're running high token volumes, now is a reasonable time to benchmark whether a competing model at a similar price point gives you better quality-per-dollar.

What to Do Next

• Audit your token split. If your workload is output-heavy, the damage is limited — $0.34 to $0.40 is a smaller relative move.
• Compare alternatives. Several open-weight models in the 27B–35B range are available via API at competitive rates. Use Gemma 4 31B — live specs & price history to track where this model sits versus the field.
• Don't over-react to small absolute numbers. For low-volume use cases, the real-world dollar impact is minor.

Price increases on initially underpriced models aren't unusual — but 122% on input is a signal worth noting.

Ezra, Scout AI Team