Qwen3.6 27B: 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 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Alibabaalibaba | Healthy | 100%◆ (best) | 99.79% | 99.80% | 99.80% | 1.50s | 15.9s | 51 tps | 262K | $0.450 / $2.70 |
| Healthy | 100%◆ (best) | 93.59% | 72.54% | 72.54% | 1.11s | 26.5s | 28 tps | 262K | $0.300 / $2.00◆ (best) | |
| Healthy | 100%◆ (best) | 99.94% | 99.91%◆ (best) | 99.91%◆ (best) | 410ms | 3.77s | 61 tps◆ (best) | 262K | $0.600 / $3.60 | |
| Healthy | 100%◆ (best) | 99.24% | 97.53% | 97.53% | 363ms◆ (best) | 2.02s◆ (best) | 43 tps | 262K | $0.320 / $3.20 | |
Phalaphala | Degraded | 100%◆ (best) | 95.23% | 97.27% | 97.27% | 1.20s | 10.9s | 47 tps | 262K | $0.320 / $2.70 |
| Healthy | 100%◆ (best) | 96.60% | 95.37% | 95.37% | 2.46s | 26.0s | 21 tps | 256K | $0.325 / $3.25 | |
| Degraded | 91.80% | 92.53% | 92.80% | 92.80% | 1.39s | 23.9s | 40 tps | 262K | $0.300 / $3.20◆ (best) |
- Alibaba100%◆ (best)Healthy24h 99.79%
- 7d
- 99.80%
- 30d
- 99.80%
- p50
- 1.50s
- p99
- 15.9s
- Throughput
- 51 tps
- Context
- 262K
- $/Mtok
- $0.450 / $2.70
alibaba - Chutes100%◆ (best)Healthy24h 93.59%
- 7d
- 72.54%
- 30d
- 72.54%
- p50
- 1.11s
- p99
- 26.5s
- Throughput
- 28 tps
- Context
- 262K
- $/Mtok
- $0.300 / $2.00◆ (best)
chutes/fp8fp8 - CoreWeave100%◆ (best)Healthy24h 99.94%
- 7d
- 99.91%◆ (best)
- 30d
- 99.91%◆ (best)
- p50
- 410ms
- p99
- 3.77s
- Throughput
- 61 tps◆ (best)
- Context
- 262K
- $/Mtok
- $0.600 / $3.60
coreweave/fp8fp8 - DeepInfra100%◆ (best)Healthy24h 99.24%
- 7d
- 97.53%
- 30d
- 97.53%
- p50
- 363ms◆ (best)
- p99
- 2.02s◆ (best)
- Throughput
- 43 tps
- Context
- 262K
- $/Mtok
- $0.320 / $3.20
deepinfra/fp8fp8 - Phala100%◆ (best)Degraded24h 95.23%
- 7d
- 97.27%
- 30d
- 97.27%
- p50
- 1.20s
- p99
- 10.9s
- Throughput
- 47 tps
- Context
- 262K
- $/Mtok
- $0.320 / $2.70
phala - Venice100%◆ (best)Healthy24h 96.60%
- 7d
- 95.37%
- 30d
- 95.37%
- p50
- 2.46s
- p99
- 26.0s
- Throughput
- 21 tps
- Context
- 256K
- $/Mtok
- $0.325 / $3.25
venice/fp8fp8 - SiliconFlow91.80%Degraded24h 92.53%
- 7d
- 92.80%
- 30d
- 92.80%
- p50
- 1.39s
- p99
- 23.9s
- Throughput
- 40 tps
- Context
- 262K
- $/Mtok
- $0.300 / $3.20◆ (best)
siliconflow/fp8fp8
Measured 55s 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.