Kimi K3: 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 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Degraded | 100%◆ (best) | 97.36% | 98.38% | 98.60% | 8.44s | 30.3s | 34 tps | 1.0M | $3.00 / $15.00 | |
| Healthy | 100%◆ (best) | 99.37% | 99.45% | 99.45% | 961ms | 28.4s | 72 tps | 1.0M | $2.85 / $14.25 | |
Fireworksfireworks | Healthy | 100%◆ (best) | 99.94% | 99.61% | 99.80%◆ (best) | 1.56s | 6.14s | 28 tps | 1.0M | $3.00 / $15.00 |
Fireworksfireworks/fast | Healthy | 100%◆ (best) | 94.58% | 98.58% | 99.03% | 801ms | 7.17s | 90 tps | 1.0M | $4.50 / $22.50 |
Fireworksfireworks/us | Healthy | 100%◆ (best) | 99.36% | 99.45% | 99.21% | 3.76s | 14.1s | 46 tps | 1.0M | $4.50 / $22.50 |
| Healthy | 100%◆ (best) | 99.76% | 99.34% | 99.34% | 684ms | 4.60s | 81 tps | 250K | $3.50 / $15.00 | |
| Healthy | 100%◆ (best) | 99.90% | 99.68% | 99.13% | 962ms | 7.88s | 59 tps | 1.0M | $0.950 / $14.00 | |
| Healthy | 100%◆ (best) | 98.61% | 98.87% | 99.56% | 1.28s | 12.4s | 70 tps | 1.0M | $3.00 / $15.00 | |
| Healthy | 100%◆ (best) | 99.99% | 99.55% | 99.67% | 5.31s | 32.9s | 30 tps | 1.0M | $3.00 / $15.00 | |
| Healthy | 100%◆ (best) | 98.89% | 98.28% | 98.96% | 1.53s | 6.51s | 36 tps | 1.0M | $1.55 / $12.83 | |
Waferwafer | Healthy | 100%◆ (best) | 99.42% | 99.79%◆ (best) | 99.55% | 659ms | 13.0s | 48 tps | 1.0M | $1.00 / $14.00 |
Waferwafer/us | Healthy | 100%◆ (best) | 98.70% | 99.55% | 99.55% | 718ms | 25.3s | 54 tps | 1.0M | $2.80 / $14.00 |
| Healthy | 99.64% | 99.78% | 99.76% | 99.64% | 1.34s | 5.88s | 83 tps | 1.0M | $2.80 / $14.00 | |
| Healthy | 99.62% | 99.69% | 90.45% | 96.73% | 1.65s | 15.6s | 41 tps | 1.0M | $0.940 / $13.00◆ (best) | |
Togethertogether | Healthy | 98.43% | 99.28% | 99.40% | 99.34% | 883ms | 7.55s | 57 tps | 1.0M | $2.70 / $13.50 |
Makoramakora | Healthy | 98.31% | 95.83% | 97.78% | 97.06% | 1.02s | 24.2s | 54 tps | 1.0M | $1.53 / $12.75 |
| Healthy | 98.30% | 98.34% | 98.59% | 97.38% | 1.18s | 18.2s | 75 tps | 1.0M | $3.00 / $15.00 | |
Phalaphala | Healthy | 96.81% | 96.60% | 98.10% | 98.84% | 2.76s | 40.0s | 27 tps | 1.0M | $1.95 / $9.75 |
| No data | — | 98.53% | 99.73% | 99.73% | 667ms | 7.34s | 65 tps | 1.0M | $1.30 / $14.00 | |
Alibabaalibaba | No data | — | 98.98% | 98.91% | 98.94% | 1.79s | 62.3s | 44 tps | 1.0M | $3.45 / $17.25 |
Amazon Bedrockamazon-bedrock/us-east-2 | No data | — | 94.58% | 96.11% | 96.11% | 2.73s | 42.6s | 66 tps | 1.0M | $3.00 / $15.00 |
| No data | — | 97.29% | 94.50% | 97.42% | 2.55s | 13.4s | 24 tps | 1.0M | $3.00 / $15.00 | |
| No data | — | 70.33% | 90.30% | 90.25% | 651ms◆ (best) | 5.45s | 125 tps◆ (best) | 1.0M | $2.52 / $12.60 | |
| No data | — | — | 98.78% | 98.06% | — | — | — | 1.0M | $2.85 / $14.25 | |
DigitalOceandigitalocean | No data | — | 99.95% | 99.51% | 99.41% | 868ms | 2.93s◆ (best) | 43 tps | 1.0M | $2.55 / $12.95 |
InferenceNetinference-net | No data | — | — | — | 96.45% | — | — | — | 1.0M | $2.10 / $10.95 |
Morphmorph | No data | — | — | — | 99.41% | — | — | — | 1.0M | $2.10 / $11.74 |
Morphmorph/fast | No data | — | — | — | 97.50% | — | — | — | 1.0M | $6.00 / $22.50 |
| No data | — | — | — | 98.42% | — | — | — | 1.0M | $2.50 / $14.00 |
- BaseTen100%◆ (best)Degraded24h 97.36%
- 7d
- 98.38%
- 30d
- 98.60%
- p50
- 8.44s
- p99
- 30.3s
- Throughput
- 34 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
baseten/fp8fp8 - DeepInfra100%◆ (best)Healthy24h 99.37%
- 7d
- 99.45%
- 30d
- 99.45%
- p50
- 961ms
- p99
- 28.4s
- Throughput
- 72 tps
- Context
- 1.0M
- $/Mtok
- $2.85 / $14.25
deepinfra/mxfp4mxfp4 - Fireworks100%◆ (best)Healthy24h 99.94%
- 7d
- 99.61%
- 30d
- 99.80%◆ (best)
- p50
- 1.56s
- p99
- 6.14s
- Throughput
- 28 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
fireworks - Fireworks100%◆ (best)Healthy24h 94.58%
- 7d
- 98.58%
- 30d
- 99.03%
- p50
- 801ms
- p99
- 7.17s
- Throughput
- 90 tps
- Context
- 1.0M
- $/Mtok
- $4.50 / $22.50
fireworks/fast - Fireworks100%◆ (best)Healthy24h 99.36%
- 7d
- 99.45%
- 30d
- 99.21%
- p50
- 3.76s
- p99
- 14.1s
- Throughput
- 46 tps
- Context
- 1.0M
- $/Mtok
- $4.50 / $22.50
fireworks/us - InferenceNet100%◆ (best)Healthy24h 99.76%
- 7d
- 99.34%
- 30d
- 99.34%
- p50
- 684ms
- p99
- 4.60s
- Throughput
- 81 tps
- Context
- 250K
- $/Mtok
- $3.50 / $15.00
inference-net/fastfp4 - InferenceNet100%◆ (best)Healthy24h 99.90%
- 7d
- 99.68%
- 30d
- 99.13%
- p50
- 962ms
- p99
- 7.88s
- Throughput
- 59 tps
- Context
- 1.0M
- $/Mtok
- $0.950 / $14.00
inference-net/fp4fp4 - Modal100%◆ (best)Healthy24h 98.61%
- 7d
- 98.87%
- 30d
- 99.56%
- p50
- 1.28s
- p99
- 12.4s
- Throughput
- 70 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
modal/mxfp4mxfp4 - Moonshot AI100%◆ (best)Healthy24h 99.99%
- 7d
- 99.55%
- 30d
- 99.67%
- p50
- 5.31s
- p99
- 32.9s
- Throughput
- 30 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
moonshotai/mxfp4mxfp4 - Morph100%◆ (best)Healthy24h 98.89%
- 7d
- 98.28%
- 30d
- 98.96%
- p50
- 1.53s
- p99
- 6.51s
- Throughput
- 36 tps
- Context
- 1.0M
- $/Mtok
- $1.55 / $12.83
morph/fp8fp8 - Wafer100%◆ (best)Healthy24h 99.42%
- 7d
- 99.79%◆ (best)
- 30d
- 99.55%
- p50
- 659ms
- p99
- 13.0s
- Throughput
- 48 tps
- Context
- 1.0M
- $/Mtok
- $1.00 / $14.00
wafer - Wafer100%◆ (best)Healthy24h 98.70%
- 7d
- 99.55%
- 30d
- 99.55%
- p50
- 718ms
- p99
- 25.3s
- Throughput
- 54 tps
- Context
- 1.0M
- $/Mtok
- $2.80 / $14.00
wafer/us - Sail Research99.64%Healthy24h 99.78%
- 7d
- 99.76%
- 30d
- 99.64%
- p50
- 1.34s
- p99
- 5.88s
- Throughput
- 83 tps
- Context
- 1.0M
- $/Mtok
- $2.80 / $14.00
sail-research/fp4fp4 - Relace99.62%Healthy24h 99.69%
- 7d
- 90.45%
- 30d
- 96.73%
- p50
- 1.65s
- p99
- 15.6s
- Throughput
- 41 tps
- Context
- 1.0M
- $/Mtok
- $0.940 / $13.00◆ (best)
relace/fp4fp4 - Together98.43%Healthy24h 99.28%
- 7d
- 99.40%
- 30d
- 99.34%
- p50
- 883ms
- p99
- 7.55s
- Throughput
- 57 tps
- Context
- 1.0M
- $/Mtok
- $2.70 / $13.50
together - Makora98.31%Healthy24h 95.83%
- 7d
- 97.78%
- 30d
- 97.06%
- p50
- 1.02s
- p99
- 24.2s
- Throughput
- 54 tps
- Context
- 1.0M
- $/Mtok
- $1.53 / $12.75
makora - Parasail98.30%Healthy24h 98.34%
- 7d
- 98.59%
- 30d
- 97.38%
- p50
- 1.18s
- p99
- 18.2s
- Throughput
- 75 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
parasail/fp4fp4 - Phala96.81%Healthy24h 96.60%
- 7d
- 98.10%
- 30d
- 98.84%
- p50
- 2.76s
- p99
- 40.0s
- Throughput
- 27 tps
- Context
- 1.0M
- $/Mtok
- $1.95 / $9.75
phala - No data24h 98.53%
- 7d
- 99.73%
- 30d
- 99.73%
- p50
- 667ms
- p99
- 7.34s
- Throughput
- 65 tps
- Context
- 1.0M
- $/Mtok
- $1.30 / $14.00
akashml/fp4fp4 - No data24h 98.98%
- 7d
- 98.91%
- 30d
- 98.94%
- p50
- 1.79s
- p99
- 62.3s
- Throughput
- 44 tps
- Context
- 1.0M
- $/Mtok
- $3.45 / $17.25
alibaba - No data24h 94.58%
- 7d
- 96.11%
- 30d
- 96.11%
- p50
- 2.73s
- p99
- 42.6s
- Throughput
- 66 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
amazon-bedrock/us-east-2 - No data24h 97.29%
- 7d
- 94.50%
- 30d
- 97.42%
- p50
- 2.55s
- p99
- 13.4s
- Throughput
- 24 tps
- Context
- 1.0M
- $/Mtok
- $3.00 / $15.00
chutes/mxfp4mxfp4 - No data24h 70.33%
- 7d
- 90.30%
- 30d
- 90.25%
- p50
- 651ms◆ (best)
- p99
- 5.45s
- Throughput
- 125 tps◆ (best)
- Context
- 1.0M
- $/Mtok
- $2.52 / $12.60
decart/mxfp4mxfp4 - No data24h —
- 7d
- 98.78%
- 30d
- 98.06%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 1.0M
- $/Mtok
- $2.85 / $14.25
deepinfra/bf16bf16 - No data24h 99.95%
- 7d
- 99.51%
- 30d
- 99.41%
- p50
- 868ms
- p99
- 2.93s◆ (best)
- Throughput
- 43 tps
- Context
- 1.0M
- $/Mtok
- $2.55 / $12.95
digitalocean - No data24h —
- 7d
- —
- 30d
- 96.45%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 1.0M
- $/Mtok
- $2.10 / $10.95
inference-net - No data24h —
- 7d
- —
- 30d
- 99.41%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 1.0M
- $/Mtok
- $2.10 / $11.74
morph - No data24h —
- 7d
- —
- 30d
- 97.50%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 1.0M
- $/Mtok
- $6.00 / $22.50
morph/fast - No data24h —
- 7d
- —
- 30d
- 98.42%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 1.0M
- $/Mtok
- $2.50 / $14.00
morph/fp4fp4
Measured 3m 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.