DeepSeek V3.2 Exp
deepseek/deepseek-v3.2-exp
Compare with DeepSeek V3.2, DeepSeek V3, DeepSeek V3.1
CompareDeepSeek V3.2 Exp
| Provider | State | Uptime 5m | 24h | p50 latency | p99 | Throughput | Context | $/Mtok |
|---|---|---|---|---|---|---|---|---|
AtlasCloudatlas-cloud/fp8fp8 | Healthy | 100% | 99.29% | 1.55s | 9.79s | 20 tps | 164K | $0.270 / $0.410 |
SiliconFlowsiliconflow/fp8fp8 | Healthy | 100% | 99.89% | 1.64s | 10.8s | 21 tps | 164K | $0.270 / $0.410 |
Novitanovita/fp8fp8 | Degraded | 93.77% | 99.84% | 1.70s | 11.6s | 18 tps | 164K | $0.270 / $0.410 |
- AtlasCloud100%Healthy24h 99.29%
- p50
- 1.55s
- p99
- 9.79s
- Throughput
- 20 tps
- Context
- 164K
- $/Mtok
- $0.270 / $0.410
atlas-cloud/fp8fp8 - SiliconFlow100%Healthy24h 99.89%
- p50
- 1.64s
- p99
- 10.8s
- Throughput
- 21 tps
- Context
- 164K
- $/Mtok
- $0.270 / $0.410
siliconflow/fp8fp8 - Novita93.77%Degraded24h 99.84%
- p50
- 1.70s
- p99
- 11.6s
- Throughput
- 18 tps
- Context
- 164K
- $/Mtok
- $0.270 / $0.410
novita/fp8fp8
Measured 2m 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.