DeepSeek V3: 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 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Healthy | 99.54%◆ (best) | 96.18% | 95.58% | 94.77% | 635ms◆ (best) | 6.83s | 16 tps | 164K | $0.320 / $0.890 | |
StreamLakestreamlake | Healthy | 98.75% | 98.86% | 98.86%◆ (best) | 98.22%◆ (best) | 1.36s | 5.12s◆ (best) | 45 tps◆ (best) | 128K | $0.257 / $1.03◆ (best) |
| No data | — | — | — | — | — | — | — | 64K | $0.400 / $1.30 |
- DeepInfra99.54%◆ (best)Healthy24h 96.18%
- 7d
- 95.58%
- 30d
- 94.77%
- p50
- 635ms◆ (best)
- p99
- 6.83s
- Throughput
- 16 tps
- Context
- 164K
- $/Mtok
- $0.320 / $0.890
deepinfra/fp4fp4 - StreamLake98.75%Healthy24h 98.86%
- 7d
- 98.86%◆ (best)
- 30d
- 98.22%◆ (best)
- p50
- 1.36s
- p99
- 5.12s◆ (best)
- Throughput
- 45 tps◆ (best)
- Context
- 128K
- $/Mtok
- $0.257 / $1.03◆ (best)
streamlake - No data24h —
- 7d
- —
- 30d
- —
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 64K
- $/Mtok
- $0.400 / $1.30
novita/fp8fp8
Measured 4m 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.