Qwen3.5-27B
qwen/qwen3.5-27b
Compare with Qwen3.5 397B A17B, Qwen3.5-35B-A3B, Qwen3.5-122B-A10B
CompareQwen3.5-27B
| Provider | State | Uptime 5m | 24h | p50 latency | p99 | Throughput | Context | $/Mtok |
|---|---|---|---|---|---|---|---|---|
Alibabaalibaba | Healthy | 100% | 96.86% | 1.27s | 6.05s | 23 tps | 262K | $0.195 / $1.56 |
Novitanovita/bf16bf16 | Healthy | 100% | 99.88% | 948ms | 4.96s | 14 tps | 262K | $0.300 / $2.40 |
AtlasCloudatlas-cloud/fp8fp8 | Healthy | 99.50% | 97.68% | 1.37s | 4.67s | 22 tps | 262K | $0.270 / $2.16 |
DeepInfradeepinfra/fp8fp8 | Healthy | 99.06% | 95.63% | 377ms | 1.96s | 41 tps | 262K | $0.260 / $2.60 |
SiliconFlowsiliconflow/fp8fp8 | Healthy | 97.08% | 98.76% | 1.47s | 9.24s | 11 tps | 262K | $0.250 / $2.00 |
Phalaphala | Down | — | 91.77% | 7.18s | 11.3s | 12 tps | 262K | $0.300 / $2.40 |
- Alibaba100%Healthy24h 96.86%
- p50
- 1.27s
- p99
- 6.05s
- Throughput
- 23 tps
- Context
- 262K
- $/Mtok
- $0.195 / $1.56
alibaba - Novita100%Healthy24h 99.88%
- p50
- 948ms
- p99
- 4.96s
- Throughput
- 14 tps
- Context
- 262K
- $/Mtok
- $0.300 / $2.40
novita/bf16bf16 - AtlasCloud99.50%Healthy24h 97.68%
- p50
- 1.37s
- p99
- 4.67s
- Throughput
- 22 tps
- Context
- 262K
- $/Mtok
- $0.270 / $2.16
atlas-cloud/fp8fp8 - DeepInfra99.06%Healthy24h 95.63%
- p50
- 377ms
- p99
- 1.96s
- Throughput
- 41 tps
- Context
- 262K
- $/Mtok
- $0.260 / $2.60
deepinfra/fp8fp8 - SiliconFlow97.08%Healthy24h 98.76%
- p50
- 1.47s
- p99
- 9.24s
- Throughput
- 11 tps
- Context
- 262K
- $/Mtok
- $0.250 / $2.00
siliconflow/fp8fp8 - Phala—Down24h 91.77%
- p50
- 7.18s
- p99
- 11.3s
- Throughput
- 12 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 and p99 are OpenRouter's rolling 30-minute windows.
Throughput
Median tokens per second.