modelbenchmark.io

Gemma 4 31B IT

Google · gemma-4-31b-it

Compare

Largest Gemma 4 instruction model for open, self-hosted chat and reasoning

Specification

google listing

Context
262K
Max output
33K
Released
2026-04-02
Knowledge cutoff
2025-01
Retires
2026-09-15
Open weights
yes
Input
text, image
Output
text

Price

US dollars per million tokens · google list price

Input
$0.38
Output
$1.15
Cache read
$0.05

Available from 51 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
googlevendor————262K33K—
amdfreefreefree—262K33K—
bothub · freefreefree——262K33K—
geminifreefree——262K33K—
nvidiafreefree——256K16K—
openrouter · freefreefree——262K33K—
requestyfreefree——262K8K—
unorouter · freefreefree——262K33K—
fastrouter$0.10$0.30——262K33K—
kilo$0.10$0.34$0.10—262K16K—
llmgateway$0.10$0.25$0.01—262K33K—
llmgateway-providers · consensusprotocol$0.10$0.25$0.01—262K33K—
nano-gpt · thinking$0.10$0.34$0.10—262K131K—
wandb$0.10$0.34$0.10—262K262K—
wandb · wandb$0.10$0.34——262K262K—
llmgateway-providers · runware$0.102$0.297$0.012—262K66K—
kenari$0.1029$0.4057$0.0411—262K33K—
lilac$0.11$0.35——262K262K—
orcarouter$0.13$0.38$0.02—262K33K—
abacus$0.14$0.40——262K131K—
aihubmix · aihubmix$0.14$0.40——262K131K—
crusoe$0.14$0.40$0.14—262K33K—
friendli$0.14$0.40——262K33K—
friendliai · friendliai$0.14$0.40——262K262K—
huggingface$0.14$0.40——262K33K—
llmgateway-providers
$0.14
$0.20
$0.40——262K33K—
merge-gateway$0.14$0.40——262K66K—
nano-gpt$0.14$0.40$0.05—262K131K—
novita$0.14$0.40——262K131K—
novita-ai$0.14$0.40——262K131K—
tensormesh · tensormesh$0.14$0.56free—33K33K—
vercel$0.14$0.40——262K131K—
libertai · libertai$0.15$0.40——262K262K—
nano-gpt · tee$0.15$0.46$0.075—262K262K—
digitalocean$0.18$0.50$0.036—256K8K—
aiand$0.20$0.50$0.05—262K33K—
aiand · aiand$0.20$0.50$0.05—262K33K—
deepinfra
$0.20
$0.13
$0.40
$0.38
——262K
33K
262K
—
openrouter$0.20$0.40——262K16K—
cortecs$0.223$0.39——262K262K—
infomaniak$0.25$0.50——100K33K—
modular$0.25$0.65$0.08—262K33K—
berget$0.2804$0.5607——128K8K—
sambanova$0.38$1.15——131K131K—
sambanova · sambanova$0.38$1.15——131K131K—
together_ai$0.39$0.97——262K262K2026-09-15
sail · sail$0.40$0.60$0.20—256K256K—
opper$0.4649$2.4406——256K8K—
siliconflow$0.75$1.00$0.25—262K262K—
aimlapi$1.0474$1.5764$1.0474—256K16K—
tempr————262K33K—

Price history

append-only observations · a listing writes a row only when its price moves

DateHostField$/M
2026-09-02kiloCache read$0.05
2026-10-16kiloCache read$0.10
2026-09-02llmgatewayCache read$0.012
2026-09-11llmgatewayCache read$0.01
2026-09-02nano-gpt · thinkingCache read$0.05
2026-10-19nano-gpt · thinkingCache read$0.10
2026-09-02openrouterCache read$0.05
2026-09-06openrouterCache read$0.05
2026-10-10openrouterCache read$0.05
2026-10-26openrouterCache read$0.10
2026-09-02bergetInput$0.275
2026-10-22bergetInput$0.2804
2026-09-02deepinfraInput$0.13
2026-10-02deepinfraInput$0.15
2026-10-05deepinfraInput$0.20
2026-09-02fastrouterInput$0.13
2026-10-22fastrouterInput$0.10
2026-09-02kenariInputfree
2026-10-22kenariInput$0.1029
2026-09-02kiloInput$0.09
2026-09-27kiloInput$0.08
2026-10-01kiloInput$0.09
2026-10-16kiloInput$0.10
2026-09-02llmgatewayInput$0.102
2026-09-11llmgatewayInput$0.10
2026-09-02llmgateway-providersInput$0.13
2026-09-04llmgateway-providersInput$0.39
2026-09-15llmgateway-providersInput$0.14
2026-09-24llmgateway-providersInput$0.13
2026-09-28llmgateway-providersInput$0.15
2026-10-05llmgateway-providersInput$0.20
2026-10-19llmgateway-providersInput$0.99
2026-10-20llmgateway-providersInput$0.14
2026-10-21llmgateway-providersInput$0.20
2026-10-22llmgateway-providersInput$0.20
2026-10-23llmgateway-providersInput$0.99
2026-10-24llmgateway-providersInput$0.14
2026-10-25llmgateway-providersInput$0.99
2026-10-27llmgateway-providersInput$0.20
2026-10-28llmgateway-providersInput$0.14
2026-10-29llmgateway-providersInput$0.14
2026-10-31llmgateway-providersInput$0.20
2026-11-01llmgateway-providersInput$0.14
2026-09-02nano-gptInput$0.10
2026-10-19nano-gptInput$0.14
2026-09-02openrouterInput$0.09
2026-09-06openrouterInput$0.09
2026-10-10openrouterInput$0.09
2026-10-15openrouterInput$0.20
2026-10-26openrouterInput$0.10
2026-10-27openrouterInput$0.20
2026-09-02openrouter · freeInputfree
2026-09-06openrouter · freeInputfree
2026-10-10openrouter · freeInputfree
2026-09-02siliconflowInput$0.13
2026-10-22siliconflowInput$0.75
2026-09-02together_aiInput$0.39
2026-09-24together_aiInput$0.39
2026-09-02bergetOutput$0.55
2026-10-22bergetOutput$0.5607
2026-09-02deepinfraOutput$0.38
2026-10-02deepinfraOutput$0.40
2026-09-02fastrouterOutput$0.38
2026-10-22fastrouterOutput$0.30
2026-09-02kenariOutputfree
2026-10-22kenariOutput$0.4057
2026-09-02kiloOutput$0.34
2026-09-27kiloOutput$0.30
2026-10-01kiloOutput$0.34
2026-09-02llmgatewayOutput$0.297
2026-09-11llmgatewayOutput$0.25
2026-09-02llmgateway-providersOutput$0.38
2026-09-04llmgateway-providersOutput$0.97
2026-09-15llmgateway-providersOutput$0.40
2026-09-24llmgateway-providersOutput$0.38
2026-09-28llmgateway-providersOutput$0.40
2026-10-19llmgateway-providersOutput$1.49
2026-10-20llmgateway-providersOutput$0.40
2026-10-23llmgateway-providersOutput$1.49
2026-10-24llmgateway-providersOutput$0.40
2026-10-25llmgateway-providersOutput$1.49
2026-10-27llmgateway-providersOutput$0.40
2026-10-29llmgateway-providersOutput$0.40
2026-09-02nano-gptOutput$0.45
2026-10-19nano-gptOutput$0.40
2026-09-02nano-gpt · thinkingOutput$0.35
2026-10-19nano-gpt · thinkingOutput$0.34
2026-09-02openrouterOutput$0.34
2026-09-06openrouterOutput$0.34
2026-10-10openrouterOutput$0.34
2026-10-15openrouterOutput$0.40
2026-10-26openrouterOutput$0.34
2026-10-27openrouterOutput$0.40
2026-09-02openrouter · freeOutputfree
2026-09-06openrouter · freeOutputfree
2026-10-10openrouter · freeOutputfree
2026-09-02siliconflowOutput$0.40
2026-10-22siliconflowOutput$1.00
2026-09-02together_aiOutput$0.97
2026-09-24together_aiOutput$0.97
2026-10-13aimlapiCache read$1.3616
2026-10-23aimlapiCache read$1.0474
2026-10-13aimlapiInput$1.3616
2026-10-23aimlapiInput$1.0474
2026-10-13aimlapiOutput$2.0493
2026-10-23aimlapiOutput$1.5764

Other listings of this model

same model, different key

More from Google

most-hosted first