AllocateGemma 4 31B-it FP8 vs MiniMax M2.7 FP4
On provider list prices, Gemma 4 31B-it FP8 costs $0.39 per million input tokens against $0.30 for MiniMax M2.7 FP4: effectively level. Output is $0.97 against $1.20 (1.2x).
Gemma 4 31B-it FP8M MiniMax M2.7 FP4Specifications and provider list prices from the Allocate catalog, checked 2026-07-21.
What the numbers say
Take 1,000,000 requests a month at 1,200 input and 350 output tokens each. That workload costs $780 a month on MiniMax M2.7 FP4 and $807.50 on Gemma 4 31B-it FP8 at list: a gap of $27.50.
Gemma 4 31B-it FP8 reads 256K tokens per request against 192K for MiniMax M2.7 FP4, 1.3x the window. That decides which one can take whole documents without splitting them.
Choose Gemma 4 31B-it FP8 for
- The longer context window (256K vs 192K tokens)
- Fine-tuning under a permissive license (Apache 2.0)
Choose MiniMax M2.7 FP4 for
- The lower list price ($0.30 in / $1.20 out per M tokens)
- Published cached-input pricing ($0.06 per M tokens)
Common questions
Which is cheaper, Gemma 4 31B-it FP8 or MiniMax M2.7 FP4?
MiniMax M2.7 FP4, on this workload shape. At list prices it is $0.30/$1.20 per million tokens in and out against $0.39/$0.97 for Gemma 4 31B-it FP8. Billed on Allocate: $0.32/$1.28 against $0.42/$1.04.
Which has the bigger context window?
Gemma 4 31B-it FP8: 262,144 tokens (256K) against 196,608 (192K) for MiniMax M2.7 FP4.
Can I fine-tune Gemma 4 31B-it FP8 or MiniMax M2.7 FP4?
Both publish open weights (Gemma 4 31B-it FP8: Apache 2.0; MiniMax M2.7 FP4: Not listed), so both can be fine-tuned. On Allocate the trained weights stay inside your boundary and belong to you.
Related comparisons
Run the numbers on your workload
Or do not choose. On Allocate a route name is the contract: point yours at one model today, swap to the other tomorrow, and compare them on your live traffic with per-token metering.