MiniMax M3: providers compared

Every endpoint serving this model, side by side. ◆ marks the best value in each column — per model, never fleet-wide.

Current state

	State
AtlasCloudatlas-cloud/fp8fp8	Healthy	100%◆ (best)	99.68%	99.53%	99.53%	1.31s	40.3s	75 tps	524K	$0.300 / $1.20
CoreWeavecoreweave/fp4fp4	Healthy	100%◆ (best)	99.68%	97.50%	97.50%	529ms◆ (best)	2.56s◆ (best)	116 tps	262K	$0.230 / $0.960◆ (best)
DeepInfradeepinfra/fp8fp8	Healthy	100%◆ (best)	98.05%	98.27%	98.27%	552ms	6.92s	31 tps	524K	$0.280 / $1.10
GMICloudgmicloud/fp8fp8	Healthy	100%◆ (best)	99.50%	99.35%	99.35%	3.29s	8.36s	53 tps	1.0M	$0.240 / $0.960
ModelRunmodelrun/fp4fp4	Healthy	100%◆ (best)	99.80%	99.77%	99.77%	687ms	18.7s	119 tps◆ (best)	1.0M	$0.750 / $3.00
SambaNovasambanova	Healthy	100%◆ (best)	98.12%	98.63%	98.63%	2.93s	13.4s	117 tps	1.0M	$0.600 / $2.40
Venicevenice/fp8fp8	Healthy	100%◆ (best)	99.88%	99.82%◆ (best)	99.82%◆ (best)	1.56s	12.8s	85 tps	524K	$0.300 / $1.20
Novitanovita/fp8fp8	Healthy	99.72%	99.60%	99.66%	99.66%	2.21s	22.3s	49 tps	1.0M	$0.300 / $1.20
Minimaxminimax/fp8fp8	Healthy	99.34%	99.51%	99.64%	99.64%	1.21s	13.0s	69 tps	524K	$0.300 / $1.20
Togethertogether	Healthy	98.98%	99.63%	99.76%	99.76%	908ms	9.27s	60 tps	524K	$0.300 / $1.20
Parasailparasail/fp8fp8	Healthy	98.93%	98.17%	97.90%	97.90%	1.60s	26.4s	37 tps	1.0M	$0.300 / $1.20
StreamLakestreamlake/fp8fp8	Degraded	94.00%	98.92%	98.47%	98.47%	1.57s	7.34s	72 tps	1.0M	$0.300 / $1.20
Morphmorph/fp4fp4	Degraded	92.52%	96.43%	97.48%	97.48%	2.31s	21.8s	21 tps	256K	$0.255 / $1.02

AtlasCloud100%◆ (best)
Healthy24h 99.68%
7d
99.53%
30d
99.53%
p50
1.31s
p99
40.3s
Throughput
75 tps
Context
524K
$/Mtok
$0.300 / $1.20
atlas-cloud/fp8fp8
CoreWeave100%◆ (best)
Healthy24h 99.68%
7d
97.50%
30d
97.50%
p50
529ms◆ (best)
p99
2.56s◆ (best)
Throughput
116 tps
Context
262K
$/Mtok
$0.230 / $0.960◆ (best)
coreweave/fp4fp4
DeepInfra100%◆ (best)
Healthy24h 98.05%
7d
98.27%
30d
98.27%
p50
552ms
p99
6.92s
Throughput
31 tps
Context
524K
$/Mtok
$0.280 / $1.10
deepinfra/fp8fp8
GMICloud100%◆ (best)
Healthy24h 99.50%
7d
99.35%
30d
99.35%
p50
3.29s
p99
8.36s
Throughput
53 tps
Context
1.0M
$/Mtok
$0.240 / $0.960
gmicloud/fp8fp8
ModelRun100%◆ (best)
Healthy24h 99.80%
7d
99.77%
30d
99.77%
p50
687ms
p99
18.7s
Throughput
119 tps◆ (best)
Context
1.0M
$/Mtok
$0.750 / $3.00
modelrun/fp4fp4
SambaNova100%◆ (best)
Healthy24h 98.12%
7d
98.63%
30d
98.63%
p50
2.93s
p99
13.4s
Throughput
117 tps
Context
1.0M
$/Mtok
$0.600 / $2.40
sambanova
Venice100%◆ (best)
Healthy24h 99.88%
7d
99.82%◆ (best)
30d
99.82%◆ (best)
p50
1.56s
p99
12.8s
Throughput
85 tps
Context
524K
$/Mtok
$0.300 / $1.20
venice/fp8fp8
Novita99.72%
Healthy24h 99.60%
7d
99.66%
30d
99.66%
p50
2.21s
p99
22.3s
Throughput
49 tps
Context
1.0M
$/Mtok
$0.300 / $1.20
novita/fp8fp8
Minimax99.34%
Healthy24h 99.51%
7d
99.64%
30d
99.64%
p50
1.21s
p99
13.0s
Throughput
69 tps
Context
524K
$/Mtok
$0.300 / $1.20
minimax/fp8fp8
Together98.98%
Healthy24h 99.63%
7d
99.76%
30d
99.76%
p50
908ms
p99
9.27s
Throughput
60 tps
Context
524K
$/Mtok
$0.300 / $1.20
together
Parasail98.93%
Healthy24h 98.17%
7d
97.90%
30d
97.90%
p50
1.60s
p99
26.4s
Throughput
37 tps
Context
1.0M
$/Mtok
$0.300 / $1.20
parasail/fp8fp8
StreamLake94.00%
Degraded24h 98.92%
7d
98.47%
30d
98.47%
p50
1.57s
p99
7.34s
Throughput
72 tps
Context
1.0M
$/Mtok
$0.300 / $1.20
streamlake/fp8fp8
Morph92.52%
Degraded24h 96.43%
7d
97.48%
30d
97.48%
p50
2.31s
p99
21.8s
Throughput
21 tps
Context
256K
$/Mtok
$0.255 / $1.02
morph/fp4fp4

Measured 4m ago. Latency and throughput are OpenRouter's rolling 30-minute windows. ◆ marks the best value in each column among these endpoints.

Price vs latency

One dot per endpoint at the latest round · further left = faster, lower = cheaper · hover a dot for its tag

History

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, averaged over the hour.

Throughput

Median tokens per second.