Nemotron 3.5 Super VL: 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 | 100%◆ (best) | 93.50% | 87.91% | 87.91% | 670ms | 3.97s | 376 tps | 1.0M | $0.150 / $0.750 | |
| Healthy | 100%◆ (best) | 95.66% | 97.67% | 97.67% | 310ms◆ (best) | 8.02s | 99 tps | 262K | $0.100 / $0.400◆ (best) | |
Togethertogether | Healthy | 100%◆ (best) | 98.98% | 99.53% | 99.53% | 377ms | 2.82s | 273 tps | 262K | $0.150 / $0.500 |
| No data | — | 99.74% | 100%◆ (best) | 100%◆ (best) | 426ms | 2.42s◆ (best) | 431 tps◆ (best) | 262K | $0.200 / $1.00 | |
| No data | — | 100% | 100%◆ (best) | 100%◆ (best) | 518ms | 4.59s | 422 tps | 262K | $0.250 / $1.25 |
- CoreWeave100%◆ (best)Healthy24h 93.50%
- 7d
- 87.91%
- 30d
- 87.91%
- p50
- 670ms
- p99
- 3.97s
- Throughput
- 376 tps
- Context
- 1.0M
- $/Mtok
- $0.150 / $0.750
coreweave/nvfp4nvfp4 - Crusoe100%◆ (best)Healthy24h 95.66%
- 7d
- 97.67%
- 30d
- 97.67%
- p50
- 310ms◆ (best)
- p99
- 8.02s
- Throughput
- 99 tps
- Context
- 262K
- $/Mtok
- $0.100 / $0.400◆ (best)
crusoe/bf16bf16 - Together100%◆ (best)Healthy24h 98.98%
- 7d
- 99.53%
- 30d
- 99.53%
- p50
- 377ms
- p99
- 2.82s
- Throughput
- 273 tps
- Context
- 262K
- $/Mtok
- $0.150 / $0.500
together - No data24h 99.74%
- 7d
- 100%◆ (best)
- 30d
- 100%◆ (best)
- p50
- 426ms
- p99
- 2.42s◆ (best)
- Throughput
- 431 tps◆ (best)
- Context
- 262K
- $/Mtok
- $0.200 / $1.00
deepinfra/fp4fp4 - No data24h 100%
- 7d
- 100%◆ (best)
- 30d
- 100%◆ (best)
- p50
- 518ms
- p99
- 4.59s
- Throughput
- 422 tps
- Context
- 262K
- $/Mtok
- $0.250 / $1.25
venice/fp4fp4
Measured 5m 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, averaged over the hour.
Throughput
Median tokens per second.