DeepSeek V3
deepseek/deepseek-chat
Compare with DeepSeek V4 Flash 0731, DeepSeek V4 Flash 0423, DeepSeek V4 Pro 0423
CompareDeepSeek V3
| State | ||||||||
|---|---|---|---|---|---|---|---|---|
| Degraded | 100% | 90.70% | 1.76s | 7.96s | 12 tps | 164K | $0.320 / $0.890 | |
StreamLakestreamlake | Healthy | 99.78% | 99.01% | 1.40s | 5.38s | 28 tps | 128K | $0.257 / $1.03 |
| Healthy | 99.71% | 97.01% | 1.07s | 3.64s | 22 tps | 64K | $0.400 / $1.30 |
- DeepInfra100%Degraded24h 90.70%
- p50
- 1.76s
- p99
- 7.96s
- Throughput
- 12 tps
- Context
- 164K
- $/Mtok
- $0.320 / $0.890
deepinfra/fp4fp4 - StreamLake99.78%Healthy24h 99.01%
- p50
- 1.40s
- p99
- 5.38s
- Throughput
- 28 tps
- Context
- 128K
- $/Mtok
- $0.257 / $1.03
streamlake - Novita99.71%Healthy24h 97.01%
- p50
- 1.07s
- p99
- 3.64s
- Throughput
- 22 tps
- Context
- 64K
- $/Mtok
- $0.400 / $1.30
novita/fp8fp8
Measured 4m 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.