Qwen3.5-27B
qwen/qwen3.5-27b
Compare with Qwen3.5 397B A17B, Qwen3.5-35B-A3B, Qwen3.5-122B-A10B
CompareQwen3.5-27B
| State | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Alibabaalibaba | Healthy | 100% | 96.99% | 98.17% | 98.17% | 1.03s | 12.2s | 23 tps | 262K | $0.195 / $1.56 |
| Healthy | 100% | 99.12% | 98.97% | 98.97% | 2.97s | 9.19s | 29 tps | 262K | $0.270 / $2.16 | |
| Healthy | 100% | 98.81% | 97.87% | 97.87% | 380ms | 5.25s | 38 tps | 262K | $0.260 / $2.60 | |
| Healthy | 100% | 99.70% | 99.79% | 99.79% | 1.11s | 5.95s | 12 tps | 262K | $0.300 / $2.40 | |
| Healthy | 100% | 98.83% | 98.97% | 98.97% | 1.95s | 27.7s | 11 tps | 262K | $0.250 / $2.00 | |
Phalaphala | Degraded | — | 67.95% | 62.74% | 62.74% | 1.28s | 11.7s | 11 tps | 262K | $0.300 / $2.40 |
- Alibaba100%Healthy24h 96.99%
- 7d
- 98.17%
- 30d
- 98.17%
- p50
- 1.03s
- p99
- 12.2s
- Throughput
- 23 tps
- Context
- 262K
- $/Mtok
- $0.195 / $1.56
alibaba - AtlasCloud100%Healthy24h 99.12%
- 7d
- 98.97%
- 30d
- 98.97%
- p50
- 2.97s
- p99
- 9.19s
- Throughput
- 29 tps
- Context
- 262K
- $/Mtok
- $0.270 / $2.16
atlas-cloud/fp8fp8 - DeepInfra100%Healthy24h 98.81%
- 7d
- 97.87%
- 30d
- 97.87%
- p50
- 380ms
- p99
- 5.25s
- Throughput
- 38 tps
- Context
- 262K
- $/Mtok
- $0.260 / $2.60
deepinfra/fp8fp8 - Novita100%Healthy24h 99.70%
- 7d
- 99.79%
- 30d
- 99.79%
- p50
- 1.11s
- p99
- 5.95s
- Throughput
- 12 tps
- Context
- 262K
- $/Mtok
- $0.300 / $2.40
novita/bf16bf16 - SiliconFlow100%Healthy24h 98.83%
- 7d
- 98.97%
- 30d
- 98.97%
- p50
- 1.95s
- p99
- 27.7s
- Throughput
- 11 tps
- Context
- 262K
- $/Mtok
- $0.250 / $2.00
siliconflow/fp8fp8 - Degraded24h 67.95%
- 7d
- 62.74%
- 30d
- 62.74%
- p50
- 1.28s
- p99
- 11.7s
- Throughput
- 11 tps
- Context
- 262K
- $/Mtok
- $0.300 / $2.40
phala
Measured 3m ago. Latency and throughput are OpenRouter's rolling 30-minute windows.
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.