Gemma 4 31B

google/gemma-4-31b-it

Compare with Gemma 4 26B A4B , Gemma 4 26B A4B (free), Gemma 4 31B (free)

CompareGemma 4 31B

Provider	State	Uptime 5m	24h	p50 latency	p99	Throughput	Context	$/Mtok
Cerebrascerebras/fp16fp16	Healthy	100%	100%	453ms	6.15s	97 tps	131K	$0.990 / $1.49
DeepInfradeepinfra/ultrafp8	Healthy	100%	75.32%	2.42s	8.70s	22 tps	131K	$0.270 / $0.760
Friendlifriendli	Healthy	99.94%	99.80%	7.88s	13.3s	18 tps	262K	$0.140 / $0.400
DeepInfradeepinfra/turbofp4	Healthy	99.93%	99.11%	557ms	3.80s	25 tps	262K	$0.090 / $0.340
ModelRunmodelrun/fp4fp4	Healthy	99.83%	99.92%	193ms	4.09s	74 tps	262K	$0.750 / $1.00
Crusoecrusoe	Healthy	99.22%	99.00%	863ms	17.0s	19 tps	262K	$0.140 / $0.400
Venicevenice/bf16bf16	Healthy	98.93%	99.63%	1.10s	17.4s	33 tps	256K	$0.120 / $0.360
OpenInferenceopen-inference/bf16bf16	Degraded	98.88%	95.17%	2.04s	41.8s	10 tps	262K	$0.080 / $0.350
CoreWeavecoreweave/bf16bf16	Healthy	98.09%	98.15%	803ms	25.1s	27 tps	262K	$0.100 / $0.340
Parasailparasail/fp8fp8	Healthy	97.97%	99.11%	1.02s	7.55s	24 tps	262K	$0.150 / $0.400
Togethertogether	Healthy	97.83%	92.16%	806ms	12.2s	28 tps	262K	$0.390 / $0.970
SiliconFlowsiliconflow/fp8fp8	Healthy	95.62%	95.85%	2.32s	30.4s	16 tps	262K	$0.130 / $0.400
Phalaphala	Down	85.71%	95.41%	3.81s	168s	18 tps	262K	$0.150 / $0.460
Novitanovita/bf16bf16	Down	30.63%	66.35%	4.41s	73.2s	1.0 tps	262K	$0.140 / $0.400
Chuteschutes/fp4fp4	Down	9.59%	90.39%	3.13s	33.9s	9.5 tps	131K	$0.120 / $0.370
DeepInfradeepinfra/fp8fp8	Degraded	—	95.00%	2.01s	23.8s	6.0 tps	262K	$0.130 / $0.380
SambaNovasambanova	No data	—	98.21%	1.84s	8.23s	83 tps	262K	$0.380 / $1.15

Cerebras100%
Healthy24h 100%
p50
453ms
p99
6.15s
Throughput
97 tps
Context
131K
$/Mtok
$0.990 / $1.49
cerebras/fp16fp16
DeepInfra100%
Healthy24h 75.32%
p50
2.42s
p99
8.70s
Throughput
22 tps
Context
131K
$/Mtok
$0.270 / $0.760
deepinfra/ultrafp8
Friendli99.94%
Healthy24h 99.80%
p50
7.88s
p99
13.3s
Throughput
18 tps
Context
262K
$/Mtok
$0.140 / $0.400
friendli
DeepInfra99.93%
Healthy24h 99.11%
p50
557ms
p99
3.80s
Throughput
25 tps
Context
262K
$/Mtok
$0.090 / $0.340
deepinfra/turbofp4
ModelRun99.83%
Healthy24h 99.92%
p50
193ms
p99
4.09s
Throughput
74 tps
Context
262K
$/Mtok
$0.750 / $1.00
modelrun/fp4fp4
Crusoe99.22%
Healthy24h 99.00%
p50
863ms
p99
17.0s
Throughput
19 tps
Context
262K
$/Mtok
$0.140 / $0.400
crusoe
Venice98.93%
Healthy24h 99.63%
p50
1.10s
p99
17.4s
Throughput
33 tps
Context
256K
$/Mtok
$0.120 / $0.360
venice/bf16bf16
OpenInference98.88%
Degraded24h 95.17%
p50
2.04s
p99
41.8s
Throughput
10 tps
Context
262K
$/Mtok
$0.080 / $0.350
open-inference/bf16bf16
CoreWeave98.09%
Healthy24h 98.15%
p50
803ms
p99
25.1s
Throughput
27 tps
Context
262K
$/Mtok
$0.100 / $0.340
coreweave/bf16bf16
Parasail97.97%
Healthy24h 99.11%
p50
1.02s
p99
7.55s
Throughput
24 tps
Context
262K
$/Mtok
$0.150 / $0.400
parasail/fp8fp8
Together97.83%
Healthy24h 92.16%
p50
806ms
p99
12.2s
Throughput
28 tps
Context
262K
$/Mtok
$0.390 / $0.970
together
SiliconFlow95.62%
Healthy24h 95.85%
p50
2.32s
p99
30.4s
Throughput
16 tps
Context
262K
$/Mtok
$0.130 / $0.400
siliconflow/fp8fp8
Phala85.71%
Down24h 95.41%
p50
3.81s
p99
168s
Throughput
18 tps
Context
262K
$/Mtok
$0.150 / $0.460
phala
Novita30.63%
Down24h 66.35%
p50
4.41s
p99
73.2s
Throughput
1.0 tps
Context
262K
$/Mtok
$0.140 / $0.400
novita/bf16bf16
Chutes9.59%
Down24h 90.39%
p50
3.13s
p99
33.9s
Throughput
9.5 tps
Context
131K
$/Mtok
$0.120 / $0.370
chutes/fp4fp4
DeepInfra—
Degraded24h 95.00%
p50
2.01s
p99
23.8s
Throughput
6.0 tps
Context
262K
$/Mtok
$0.130 / $0.380
deepinfra/fp8fp8
SambaNova—
No data24h 98.21%
p50
1.84s
p99
8.23s
Throughput
83 tps
Context
262K
$/Mtok
$0.380 / $1.15
sambanova

Measured 5m ago. Latency and throughput are OpenRouter's rolling 30-minute windows.

History

1 endpoint reporting no data is omitted from the charts.

Availability

Axis starts at 90% so small differences stay visible; it expands to 0 when an endpoint falls below.

Time to first token

Logarithmic axis — the fleet spans two orders of magnitude. p50 and p99 are OpenRouter's rolling 30-minute windows.

Throughput

Median tokens per second.