Kimi K3
moonshotai/kimi-k3
Compare with Kimi K2.7 Code, Kimi K2.6, Kimi K2.5
CompareKimi K3
| State | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Fireworksfireworks | Healthy | 100% | 98.43% | 98.93% | 98.93% | 3.25s | 41.6s | 33 tps | 1.0M | $3.00 / $15.00 |
| Healthy | 99.85% | 99.54% | 99.65% | 99.65% | 1.60s | 17.3s | 75 tps | 1.0M | $3.00 / $15.00 | |
| Healthy | 99.67% | 97.43% | 98.49% | 98.49% | 3.46s | 92.6s | 35 tps | 975K | $2.60 / $13.00 | |
DigitalOceandigitalocean | Healthy | 99.63% | 99.45% | 99.30% | 99.30% | 2.66s | 36.6s | 24 tps | 1.0M | $2.85 / $14.25 |
| Healthy | 99.48% | 98.63% | 99.22% | 99.22% | 7.01s | 47.9s | 21 tps | 1.0M | $3.00 / $15.00 | |
| Healthy | 99.18% | 99.30% | 97.18% | 97.18% | 1.51s | 6.31s | 22 tps | 1.0M | $3.00 / $15.00 | |
| Healthy | 98.97% | 94.85% | 88.00% | 88.00% | 3.82s | 56.6s | 17 tps | 1.0M | $2.80 / $14.00 | |
Fireworksfireworks/us | Healthy | 97.79% | 92.78% | 95.41% | 95.41% | 1.40s | 29.7s | 48 tps | 1.0M | $3.30 / $16.50 |
| Healthy | 97.22% | 92.29% | 89.28% | 89.28% | 909ms | 12.0s | 63 tps | 1.0M | $2.85 / $14.25 | |
Togethertogether | Healthy | 96.44% | 98.86% | 98.74% | 98.74% | 2.34s | 29.4s | 44 tps | 1.0M | $3.00 / $15.00 |
| No data | — | 98.42% | 98.22% | 98.22% | 874ms | 13.7s | 81 tps | 1.0M | $3.00 / $15.00 | |
| No data | — | 83.96% | 82.82% | 82.82% | 4.48s | 110s | 16 tps | 1.0M | $6.00 / $22.50 | |
Phalaphala | No data | — | 96.25% | 94.92% | 94.92% | 3.46s | 108s | 4.0 tps | 1.0M | $3.00 / $15.00 |
Waferwafer | No data | — | 97.25% | 96.96% | 96.96% | — | — | — | 1.0M | $3.00 / $15.00 |
- Fireworks100%Healthy24h 98.43%
- 7d
- 98.93%
- 30d
- 98.93%
- p50
- 3.25s
- p99
- 41.6s
- Throughput
- 33 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
fireworks - Modal99.85%Healthy24h 99.54%
- 7d
- 99.65%
- 30d
- 99.65%
- p50
- 1.60s
- p99
- 17.3s
- Throughput
- 75 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
modal/mxfp4mxfp4 - Sail Research99.67%Healthy24h 97.43%
- 7d
- 98.49%
- 30d
- 98.49%
- p50
- 3.46s
- p99
- 92.6s
- Throughput
- 35 tps
- Context
- 975K
- $/Mtok
- $2.60 / $13.00
sail-research/fp4fp4 - DigitalOcean99.63%Healthy24h 99.45%
- 7d
- 99.30%
- 30d
- 99.30%
- p50
- 2.66s
- p99
- 36.6s
- Throughput
- 24 tps
- Context
- 1.0M
- $/Mtok
- $2.85 / $14.25
digitalocean - Moonshot AI99.48%Healthy24h 98.63%
- 7d
- 99.22%
- 30d
- 99.22%
- p50
- 7.01s
- p99
- 47.9s
- Throughput
- 21 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
moonshotai/mxfp4mxfp4 - Chutes99.18%Healthy24h 99.30%
- 7d
- 97.18%
- 30d
- 97.18%
- p50
- 1.51s
- p99
- 6.31s
- Throughput
- 22 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
chutes/mxfp4mxfp4 - Morph98.97%Healthy24h 94.85%
- 7d
- 88.00%
- 30d
- 88.00%
- p50
- 3.82s
- p99
- 56.6s
- Throughput
- 17 tps
- Context
- 1.0M
- $/Mtok
- $2.80 / $14.00
morph/fp4fp4 - Fireworks97.79%Healthy24h 92.78%
- 7d
- 95.41%
- 30d
- 95.41%
- p50
- 1.40s
- p99
- 29.7s
- Throughput
- 48 tps
- Context
- 1.0M
- $/Mtok
- $3.30 / $16.50
fireworks/us - DeepInfra97.22%Healthy24h 92.29%
- 7d
- 89.28%
- 30d
- 89.28%
- p50
- 909ms
- p99
- 12.0s
- Throughput
- 63 tps
- Context
- 1.0M
- $/Mtok
- $2.85 / $14.25
deepinfra/bf16bf16 - Together96.44%Healthy24h 98.86%
- 7d
- 98.74%
- 30d
- 98.74%
- p50
- 2.34s
- p99
- 29.4s
- Throughput
- 44 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
together - No data24h 98.42%
- 7d
- 98.22%
- 30d
- 98.22%
- p50
- 874ms
- p99
- 13.7s
- Throughput
- 81 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
baseten/fp8fp8 - No data24h 83.96%
- 7d
- 82.82%
- 30d
- 82.82%
- p50
- 4.48s
- p99
- 110s
- Throughput
- 16 tps
- Context
- 1.0M
- $/Mtok
- $6.00 / $22.50
morph/fastfp4 - No data24h 96.25%
- 7d
- 94.92%
- 30d
- 94.92%
- p50
- 3.46s
- p99
- 108s
- Throughput
- 4.0 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
phala - No data24h 97.25%
- 7d
- 96.96%
- 30d
- 96.96%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
wafer
Measured 3m ago. Latency and throughput are OpenRouter's rolling 30-minute windows.
History
4 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.