Kimi K2.7 Code: 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
CoreWeavecoreweave/int4int4	Healthy	100%◆ (best)	99.96%	99.92%	99.92%	747ms	14.7s	150 tps	262K	$0.710 / $3.50
Moonshot AImoonshotai/highspeedint4	Healthy	100%◆ (best)	99.85%	99.91%	99.91%	1.96s	7.72s	98 tps	262K	$1.90 / $8.00
Moonshot AImoonshotai/int4int4	Healthy	100%◆ (best)	99.88%	99.71%	99.71%	1.77s	9.09s	43 tps	262K	$0.950 / $4.00
Novitanovita/int4int4	Healthy	100%◆ (best)	99.81%	99.25%	99.25%	1.62s	10.5s	42 tps	262K	$0.912 / $3.84
Parasailparasail/int4int4	Healthy	100%◆ (best)	99.13%	99.32%	99.32%	1.36s	8.56s	46 tps	262K	$0.760 / $3.50
Inceptroninceptron/int4int4	Healthy	99.40%	99.84%	99.82%	99.82%	860ms	9.02s	70 tps	262K	$0.670 / $3.40◆ (best)
ModelRunmodelrun/fp4fp4	Degraded	97.50%	44.72%	84.20%	84.20%	964ms	126s	152 tps	262K	$0.850 / $3.75
Alibabaalibaba/fp8fp8	No data	—	99.79%	100%◆ (best)	100%◆ (best)	3.75s	9.74s	33 tps	262K	$0.950 / $4.00
Ambientambient	No data	—	99.65%	100%◆ (best)	100%◆ (best)	579ms◆ (best)	3.27s◆ (best)	86 tps	262K	$0.690 / $3.49
AtlasCloudatlas-cloud/int4int4	No data	—	—	100%◆ (best)	100%◆ (best)	—	—	—	262K	$0.950 / $4.00
Cloudflarecloudflare	No data	—	99.97%	100%◆ (best)	100%◆ (best)	861ms	9.50s	46 tps	262K	$0.950 / $4.00
DeepInfradeepinfra/fp4fp4	No data	—	99.96%	99.49%	99.49%	781ms	3.64s	57 tps	262K	$0.680 / $3.40
GMICloudgmicloud/fp8fp8	No data	—	96.49%	—	—	2.94s	25.6s	37 tps	262K	$0.950 / $4.00
SiliconFlowsiliconflow/fp8fp8	No data	—	99.94%	99.99%	99.99%	959ms	4.45s	29 tps	262K	$0.859 / $3.80
Togethertogether	No data	—	19.54%	79.02%	79.02%	893ms	94.4s	205 tps◆ (best)	262K	$0.950 / $4.00
Venicevenice/int4int4	Down	—	97.01%	97.42%	97.42%	861ms	49.1s	28 tps	256K	$0.750 / $3.50

CoreWeave100%◆ (best)
Healthy24h 99.96%
7d
99.92%
30d
99.92%
p50
747ms
p99
14.7s
Throughput
150 tps
Context
262K
$/Mtok
$0.710 / $3.50
coreweave/int4int4
Moonshot AI100%◆ (best)
Healthy24h 99.85%
7d
99.91%
30d
99.91%
p50
1.96s
p99
7.72s
Throughput
98 tps
Context
262K
$/Mtok
$1.90 / $8.00
moonshotai/highspeedint4
Moonshot AI100%◆ (best)
Healthy24h 99.88%
7d
99.71%
30d
99.71%
p50
1.77s
p99
9.09s
Throughput
43 tps
Context
262K
$/Mtok
$0.950 / $4.00
moonshotai/int4int4
Novita100%◆ (best)
Healthy24h 99.81%
7d
99.25%
30d
99.25%
p50
1.62s
p99
10.5s
Throughput
42 tps
Context
262K
$/Mtok
$0.912 / $3.84
novita/int4int4
Parasail100%◆ (best)
Healthy24h 99.13%
7d
99.32%
30d
99.32%
p50
1.36s
p99
8.56s
Throughput
46 tps
Context
262K
$/Mtok
$0.760 / $3.50
parasail/int4int4
Inceptron99.40%
Healthy24h 99.84%
7d
99.82%
30d
99.82%
p50
860ms
p99
9.02s
Throughput
70 tps
Context
262K
$/Mtok
$0.670 / $3.40◆ (best)
inceptron/int4int4
ModelRun97.50%
Degraded24h 44.72%
7d
84.20%
30d
84.20%
p50
964ms
p99
126s
Throughput
152 tps
Context
262K
$/Mtok
$0.850 / $3.75
modelrun/fp4fp4
Alibaba—
No data24h 99.79%
7d
100%◆ (best)
30d
100%◆ (best)
p50
3.75s
p99
9.74s
Throughput
33 tps
Context
262K
$/Mtok
$0.950 / $4.00
alibaba/fp8fp8
Ambient—
No data24h 99.65%
7d
100%◆ (best)
30d
100%◆ (best)
p50
579ms◆ (best)
p99
3.27s◆ (best)
Throughput
86 tps
Context
262K
$/Mtok
$0.690 / $3.49
ambient
AtlasCloud—
No data24h —
7d
100%◆ (best)
30d
100%◆ (best)
p50
—
p99
—
Throughput
—
Context
262K
$/Mtok
$0.950 / $4.00
atlas-cloud/int4int4
Cloudflare—
No data24h 99.97%
7d
100%◆ (best)
30d
100%◆ (best)
p50
861ms
p99
9.50s
Throughput
46 tps
Context
262K
$/Mtok
$0.950 / $4.00
cloudflare
DeepInfra—
No data24h 99.96%
7d
99.49%
30d
99.49%
p50
781ms
p99
3.64s
Throughput
57 tps
Context
262K
$/Mtok
$0.680 / $3.40
deepinfra/fp4fp4
GMICloud—
No data24h 96.49%
7d
—
30d
—
p50
2.94s
p99
25.6s
Throughput
37 tps
Context
262K
$/Mtok
$0.950 / $4.00
gmicloud/fp8fp8
SiliconFlow—
No data24h 99.94%
7d
99.99%
30d
99.99%
p50
959ms
p99
4.45s
Throughput
29 tps
Context
262K
$/Mtok
$0.859 / $3.80
siliconflow/fp8fp8
Together—
No data24h 19.54%
7d
79.02%
30d
79.02%
p50
893ms
p99
94.4s
Throughput
205 tps◆ (best)
Context
262K
$/Mtok
$0.950 / $4.00
together
Venice—
Down24h 97.01%
7d
97.42%
30d
97.42%
p50
861ms
p99
49.1s
Throughput
28 tps
Context
256K
$/Mtok
$0.750 / $3.50
venice/int4int4

Measured 2m 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.

Throughput

Median tokens per second.