Google: Gemma 4 31B vs nvidia/NVIDIA-Nemotron-3-Super-120B-A12B
They cost about the same for input tokens: Google: Gemma 4 31B $0.09 per M, nvidia/NVIDIA-Nemotron-3-Super-120B-A12B $0.085 per M. Google: Gemma 4 31B is 1.2× cheaper for output tokens: Google: Gemma 4 31B costs $0.34 per M and nvidia/NVIDIA-Nemotron-3-Super-120B-A12B costs $0.40 per M. Both read up to 262,144 tokens of context. Google: Gemma 4 31B accepts image, text, video; nvidia/NVIDIA-Nemotron-3-Super-120B-A12B accepts text.
| Google: Gemma 4 31B | nvidia/NVIDIA-Nemotron-3-Super-120B-A12B | |
|---|---|---|
| Model id | google/gemma-4-31b-it | nvidia/NVIDIA-Nemotron-3-Super-120B-A12B |
| Input | $0.09 per M | $0.085 per M |
| Output | $0.34 per M | $0.40 per M |
| Context length | 262,144 tokens | 262,144 tokens |
| Accepts | image, text, video | text |
| Produces | text | text |
| Served by | deepinfra |
Prices are the providers' list prices per million tokens, as of . Both run on your own provider key through one OpenAI-compatible endpoint, with no fee on the tokens.