Gemma 3 27B: 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) | 97.73% | 97.32% | 94.34% | 1.44s | 5.66s◆ (best) | 30 tps | 98K | $0.119 / $0.200 | |
| Healthy | 99.06% | 99.11% | 99.18%◆ (best) | 99.47%◆ (best) | 2.26s | 10.8s | 18 tps | 131K | $0.080 / $0.160◆ (best) | |
| Degraded | 92.17% | 96.61% | 99.11% | 99.22% | 1.16s◆ (best) | 8.71s | 23 tps | 131K | $0.080 / $0.450◆ (best) | |
| Down | 73.93% | 87.08% | 94.87% | 89.12% | 1.55s | 33.9s | 33 tps◆ (best) | 110K | $0.100 / $0.300 | |
Phalaphala | No data | — | — | — | — | — | — | — | 262K | $0.150 / $0.460 |
- Novita99.54%◆ (best)Healthy24h 97.73%
- 7d
- 97.32%
- 30d
- 94.34%
- p50
- 1.44s
- p99
- 5.66s◆ (best)
- Throughput
- 30 tps
- Context
- 98K
- $/Mtok
- $0.119 / $0.200
novita/bf16bf16 - DeepInfra99.06%Healthy24h 99.11%
- 7d
- 99.18%◆ (best)
- 30d
- 99.47%◆ (best)
- p50
- 2.26s
- p99
- 10.8s
- Throughput
- 18 tps
- Context
- 131K
- $/Mtok
- $0.080 / $0.160◆ (best)
deepinfra/fp8fp8 - Parasail92.17%Degraded24h 96.61%
- 7d
- 99.11%
- 30d
- 99.22%
- p50
- 1.16s◆ (best)
- p99
- 8.71s
- Throughput
- 23 tps
- Context
- 131K
- $/Mtok
- $0.080 / $0.450◆ (best)
parasail/fp8fp8 - Nebius73.93%Down24h 87.08%
- 7d
- 94.87%
- 30d
- 89.12%
- p50
- 1.55s
- p99
- 33.9s
- Throughput
- 33 tps◆ (best)
- Context
- 110K
- $/Mtok
- $0.100 / $0.300
nebius/fp8fp8 - No data24h —
- 7d
- —
- 30d
- —
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 262K
- $/Mtok
- $0.150 / $0.460
phala
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.
Throughput
Median tokens per second.