Gemma 4 31B
GoogleGemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function...
Entrée
$0.09
Sortie
$0.34
Blended
$0.15
Où acheter ce modèle
12 routes d'accès| Route | Entrée | Sortie | Blended | vs |
|---|---|---|---|---|
| $0.09 | $0.34 | $0.15 | - | |
| ● CoreWeave fp4 | $0.10 | $0.34 | $0.16 | +7% |
| ● Venice bf16 | $0.12 | $0.36 | $0.18 | +20% |
| ● Chutes fp4 | $0.12 | $0.37 | $0.18 | +20% |
| ● Crusoe | $0.14 | $0.40 | $0.21 | +40% |
| ● Friendli | $0.14 | $0.40 | $0.21 | +40% |
| $0.14 | $0.40 | $0.21 | +40% | |
| $0.15 | $0.40 | $0.21 | +40% | |
| $0.39 | $0.97 | $0.53 | +253% | |
| $0.38 | $1.15 | $0.57 | +280% | |
| ● ModelRun fp4 | $0.75 | $1.00 | $0.81 | +440% |
| ● SiliconFlow fp8 | $0.75 | $1.00 | $0.81 | +440% |
● API officielle · ● plateforme cloud · ● hébergeur d'inférence · USD par million de tokens. Même modèle, fournisseurs différents - le prix et la quantization varient.
- Ctx
- 262K
- max output
- 16K
- cache read /1M
- $0.05
- Famille
- gemma
- Modalités
- image · text · video
- Fonctions
- json · reasoning · tools
Prix live, mis à jour le 2026-09-15.