DeepSeek V3.2 Exp: 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 | 99.43%◆ (best) | 99.32% | 99.27%◆ (best) | 99.27%◆ (best) | 1.57s◆ (best) | 7.31s | 25 tps | 164K | $0.270 / $0.410◆ (best) | |
| Healthy | 98.82% | 97.03% | 96.62% | 96.62% | 1.71s | 9.93s | 29 tps◆ (best) | 164K | $0.270 / $0.410◆ (best) | |
| Degraded | 94.65% | 97.38% | 99.03% | 99.03% | 1.72s | 6.88s◆ (best) | 23 tps | 164K | $0.270 / $0.410◆ (best) |
- Novita99.43%◆ (best)Healthy24h 99.32%
- 7d
- 99.27%◆ (best)
- 30d
- 99.27%◆ (best)
- p50
- 1.57s◆ (best)
- p99
- 7.31s
- Throughput
- 25 tps
- Context
- 164K
- $/Mtok
- $0.270 / $0.410◆ (best)
novita/fp8fp8 - AtlasCloud98.82%Healthy24h 97.03%
- 7d
- 96.62%
- 30d
- 96.62%
- p50
- 1.71s
- p99
- 9.93s
- Throughput
- 29 tps◆ (best)
- Context
- 164K
- $/Mtok
- $0.270 / $0.410◆ (best)
atlas-cloud/fp8fp8 - SiliconFlow94.65%Degraded24h 97.38%
- 7d
- 99.03%
- 30d
- 99.03%
- p50
- 1.72s
- p99
- 6.88s◆ (best)
- Throughput
- 23 tps
- Context
- 164K
- $/Mtok
- $0.270 / $0.410◆ (best)
siliconflow/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.