AI Rate CardGemma 3 4B IT vs Gemma 4 31B IT: API price and specs
Gemma 4 31B IT is 72% cheaper than Gemma 3 4B IT on input price.
Gemma 4 31B IT is 30% cheaper than Gemma 3 4B IT on output price.
Gemma 4 31B IT is 64% cheaper than Gemma 3 4B IT on a 1M-in / 200k-out workload.
Gemma 4 31B IT has a 100% larger context than Gemma 3 4B IT.
| Field | Gemma 3 4B IT | Gemma 4 31B IT |
|---|---|---|
| Lowest input, per 1M | $0.04 +263.6% | $0.011 |
| Median input, per 1M | $0.05 | $0.14 +180% |
| Lowest output, per 1M | $0.05 +42.9% | $0.035 |
| Median output, per 1M | $0.10 | $0.40 +300% |
| Lowest unit price | — | — |
| Cost of 1M in + 200k out | $0.05 +177.8% | $0.018 |
| Context | 131K +50% | 262K |
| Max output | 131K +50% | 262K |
| Providers | 11 +70.3% | 37 |
| Regions | EU, US | CH, EU, GLOBAL, US |
| Accepts | image, image/gif, image/jpeg, image/png, image/webp, text | application/pdf, audio, file, image, image/gif, image/heic, image/heif, image/jpeg, image/png, image/webp, pdf, text, video, video/mp4, video/quicktime, video/webm |
| Returns | text | text |
| Type | Text models | Text models |
| Free option | yes | yes |
| Cheapest provider | amazon-bedrock | inferx |
Your own workload: how many tokens the model gets (input) and gives back (output) - for example, in a month. The "Cost" row of the table above is recalculated for these numbers, for each model.
Computed from current offers each time the page opens; the best value of a row is bold, the percent is how far behind the best.
Providers selling both
aihubmix, aiml, amazon-bedrock, deepinfra, edenai, fastrouter, huggingface, kilo, nvidia, openrouter, trustedrouter
Frequently asked questions
Which is cheaper, Gemma 3 4B IT or Gemma 4 31B IT?
Gemma 4 31B IT is 72% cheaper than Gemma 3 4B IT on input price. Gemma 4 31B IT is 30% cheaper than Gemma 3 4B IT on output price. Gemma 4 31B IT is 64% cheaper than Gemma 3 4B IT on a 1M-in / 200k-out workload.
Which has the larger context, Gemma 3 4B IT or Gemma 4 31B IT?
Gemma 3 4B IT: 131K tokens; Gemma 4 31B IT: 262K tokens.
Where can I use both Gemma 3 4B IT and Gemma 4 31B IT?
aihubmix, aiml, amazon-bedrock, deepinfra, edenai, fastrouter, huggingface, kilo, nvidia, openrouter, trustedrouter