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

	State
Cerebrascerebras/fp16fp16	Healthy	100%	100%	99.99%	99.99%	271ms	1.84s	176 tps	131K	$0.990 / $1.49
Crusoecrusoe	Healthy	100%	98.49%	98.58%	98.58%	809ms	18.6s	38 tps	262K	$0.140 / $0.400
Friendlifriendli	Healthy	100%	99.23%	99.54%	99.54%	687ms	6.16s	89 tps	262K	$0.140 / $0.400
ModelRunmodelrun/fp4fp4	Healthy	100%	99.92%	99.63%	99.63%	119ms	9.13s	83 tps	262K	$0.750 / $1.00
Novitanovita/bf16bf16	Healthy	100%	98.93%	77.53%	77.53%	915ms	3.69s	8.0 tps	262K	$0.140 / $0.400
Togethertogether	Healthy	100%	89.51%	95.19%	95.19%	341ms	14.0s	52 tps	262K	$0.390 / $0.970
Venicevenice/bf16bf16	Healthy	100%	99.72%	99.74%	99.74%	1.24s	7.24s	32 tps	256K	$0.120 / $0.360
DeepInfradeepinfra/turbofp4	Healthy	99.98%	99.84%	99.51%	99.51%	598ms	4.26s	45 tps	262K	$0.090 / $0.340
Parasailparasail/fp8fp8	Healthy	99.81%	99.47%	99.27%	99.27%	1.25s	6.15s	18 tps	262K	$0.150 / $0.400
DeepInfradeepinfra/fp8fp8	Healthy	99.80%	98.35%	90.35%	90.35%	597ms	7.09s	33 tps	262K	$0.130 / $0.380
CoreWeavecoreweave/bf16bf16	Healthy	98.93%	98.15%	98.52%	98.52%	825ms	23.8s	31 tps	262K	$0.100 / $0.340
OpenInferenceopen-inference/bf16bf16	Healthy	98.60%	97.11%	97.38%	97.38%	961ms	12.7s	19 tps	262K	$0.080 / $0.350
SiliconFlowsiliconflow/fp8fp8	Degraded	98.48%	74.86%	93.77%	93.77%	1.57s	17.6s	29 tps	262K	$0.130 / $0.400
Phalaphala	Down	0.00%	84.95%	91.39%	91.39%	—	—	—	262K	$0.150 / $0.460
Chuteschutes/fp4fp4	Down	—	86.50%	63.53%	63.53%	6.83s	149s	6.0 tps	131K	$0.120 / $0.370
DeepInfradeepinfra/ultrafp8	No data	—	71.87%	87.18%	87.18%	2.14s	4.79s	27 tps	131K	$0.270 / $0.760
SambaNovasambanova	No data	—	98.47%	96.94%	96.94%	956ms	7.70s	117 tps	131K	$0.380 / $1.15

Cerebras100%
Healthy24h 100%
7d
99.99%
30d
99.99%
p50
271ms
p99
1.84s
Throughput
176 tps
Context
131K
$/Mtok
$0.990 / $1.49
cerebras/fp16fp16
Crusoe100%
Healthy24h 98.49%
7d
98.58%
30d
98.58%
p50
809ms
p99
18.6s
Throughput
38 tps
Context
262K
$/Mtok
$0.140 / $0.400
crusoe
Friendli100%
Healthy24h 99.23%
7d
99.54%
30d
99.54%
p50
687ms
p99
6.16s
Throughput
89 tps
Context
262K
$/Mtok
$0.140 / $0.400
friendli
ModelRun100%
Healthy24h 99.92%
7d
99.63%
30d
99.63%
p50
119ms
p99
9.13s
Throughput
83 tps
Context
262K
$/Mtok
$0.750 / $1.00
modelrun/fp4fp4
Novita100%
Healthy24h 98.93%
7d
77.53%
30d
77.53%
p50
915ms
p99
3.69s
Throughput
8.0 tps
Context
262K
$/Mtok
$0.140 / $0.400
novita/bf16bf16
Together100%
Healthy24h 89.51%
7d
95.19%
30d
95.19%
p50
341ms
p99
14.0s
Throughput
52 tps
Context
262K
$/Mtok
$0.390 / $0.970
together
Venice100%
Healthy24h 99.72%
7d
99.74%
30d
99.74%
p50
1.24s
p99
7.24s
Throughput
32 tps
Context
256K
$/Mtok
$0.120 / $0.360
venice/bf16bf16
DeepInfra99.98%
Healthy24h 99.84%
7d
99.51%
30d
99.51%
p50
598ms
p99
4.26s
Throughput
45 tps
Context
262K
$/Mtok
$0.090 / $0.340
deepinfra/turbofp4
Parasail99.81%
Healthy24h 99.47%
7d
99.27%
30d
99.27%
p50
1.25s
p99
6.15s
Throughput
18 tps
Context
262K
$/Mtok
$0.150 / $0.400
parasail/fp8fp8
DeepInfra99.80%
Healthy24h 98.35%
7d
90.35%
30d
90.35%
p50
597ms
p99
7.09s
Throughput
33 tps
Context
262K
$/Mtok
$0.130 / $0.380
deepinfra/fp8fp8
CoreWeave98.93%
Healthy24h 98.15%
7d
98.52%
30d
98.52%
p50
825ms
p99
23.8s
Throughput
31 tps
Context
262K
$/Mtok
$0.100 / $0.340
coreweave/bf16bf16
OpenInference98.60%
Healthy24h 97.11%
7d
97.38%
30d
97.38%
p50
961ms
p99
12.7s
Throughput
19 tps
Context
262K
$/Mtok
$0.080 / $0.350
open-inference/bf16bf16
SiliconFlow98.48%
Degraded24h 74.86%
7d
93.77%
30d
93.77%
p50
1.57s
p99
17.6s
Throughput
29 tps
Context
262K
$/Mtok
$0.130 / $0.400
siliconflow/fp8fp8
Phala0.00%
Down24h 84.95%
7d
91.39%
30d
91.39%
p50
—
p99
—
Throughput
—
Context
262K
$/Mtok
$0.150 / $0.460
phala
Chutes—
Down24h 86.50%
7d
63.53%
30d
63.53%
p50
6.83s
p99
149s
Throughput
6.0 tps
Context
131K
$/Mtok
$0.120 / $0.370
chutes/fp4fp4
DeepInfra—
No data24h 71.87%
7d
87.18%
30d
87.18%
p50
2.14s
p99
4.79s
Throughput
27 tps
Context
131K
$/Mtok
$0.270 / $0.760
deepinfra/ultrafp8
SambaNova—
No data24h 98.47%
7d
96.94%
30d
96.94%
p50
956ms
p99
7.70s
Throughput
117 tps
Context
131K
$/Mtok
$0.380 / $1.15
sambanova

Measured 4m ago. Latency and throughput are OpenRouter's rolling 30-minute windows.

Price changes

Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 08:35 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 08:30 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 08:25 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 08:20 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 08:15 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 08:10 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 08:05 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 08:00 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:55 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:50 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:45 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:40 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:35 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:30 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:25 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:20 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:15 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:10 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:05 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 07:00 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 06:55 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 06:50 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 06:45 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 06:40 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 06:35 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 06:30 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 06:25 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 06:20 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 06:15 UTC
Gemma 4 31B Togethertogether$0.390 / $0.970 → $0.280 / $0.8602026-08-21 06:10 UTC

History

2 endpoints reporting no data are 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 is OpenRouter's rolling 30-minute window.

Throughput

Median tokens per second.