gpt-oss-120b: 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 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Amazon Bedrockamazon-bedrock | Healthy | 100%◆ (best) | 99.40% | 99.91% | 98.18% | 603ms | 17.0s | 143 tps | 131K | $0.150 / $0.600 |
| Healthy | 100%◆ (best) | 99.97% | 99.85% | 99.93% | 265ms | 5.76s | 124 tps | 128K | $0.100 / $0.500 | |
| Healthy | 100%◆ (best) | 99.98% | 99.95% | 99.95% | 243ms | 3.68s | 745 tps◆ (best) | 131K | $0.350 / $0.750 | |
| Healthy | 100%◆ (best) | 99.99% | 99.96% | 99.64% | 208ms◆ (best) | 5.45s | 131 tps | 131K | $0.050 / $0.250 | |
| Healthy | 100%◆ (best) | 99.98% | 99.96% | 99.97%◆ (best) | 561ms | 2.05s | 98 tps | 131K | $0.150 / $0.600 | |
DigitalOceandigitalocean | Healthy | 100%◆ (best) | 99.98% | 99.94% | 99.94% | 483ms | 1.94s◆ (best) | 25 tps | 128K | $0.060 / $0.420 |
| Healthy | 100%◆ (best) | 99.97% | 99.89% | 99.35% | 406ms | 1.98s | 86 tps | 131K | $0.100 / $0.750 | |
| Healthy | 99.99% | 99.95% | 99.97%◆ (best) | 99.91% | 947ms | 33.0s | 32 tps | 131K | $0.037 / $0.187 | |
| Healthy | 99.93% | 99.53% | 99.47% | 99.43% | 901ms | 49.8s | 19 tps | 131K | $0.030 / $0.180◆ (best) | |
Groqgroq | Healthy | 99.90% | 99.40% | 99.08% | 99.64% | 389ms | 2.49s | 274 tps | 131K | $0.150 / $0.600 |
| Healthy | 99.80% | 98.92% | 98.99% | 99.68% | 475ms | 25.5s | 32 tps | 131K | $0.030 / $0.170◆ (best) | |
SambaNovasambanova | Healthy | 99.71% | 99.76% | 99.39% | 96.63% | 800ms | 7.58s | 214 tps | 131K | $0.140 / $0.950 |
Togethertogether | Healthy | 98.10% | 88.62% | 92.68% | 92.63% | 261ms | 4.95s | 79 tps | 131K | $0.150 / $0.600 |
| Degraded | 93.27% | 97.50% | 99.00% | 99.07% | 1.50s | 47.5s | 25 tps | 131K | $0.045 / $0.250 | |
Phalaphala | Degraded | 91.82% | 98.96% | 99.31% | 92.31% | 962ms | 9.81s | 93 tps | 131K | $0.150 / $0.600 |
| Degraded | 90.75% | 98.46% | 99.10% | 92.16% | 845ms | 19.4s | 104 tps | 131K | $0.050 / $0.250 | |
| Down | 88.95% | 97.73% | 96.46% | 97.77% | 441ms | 11.5s | 132 tps | 131K | $0.150 / $0.600 | |
Googlegoogle-vertex/global | Down | 78.17% | 66.73% | 54.88% | 78.16% | 534ms | 7.77s | 101 tps | 131K | $0.090 / $0.360 |
Maramara | Down | 74.60% | 95.80% | 93.43% | 85.85% | 1.02s | 10.9s | 118 tps | 131K | $0.150 / $0.750 |
| Down | 64.61% | 97.18% | 98.76% | 99.05% | 13.1s | 41.6s | 13 tps | 131K | $0.037 / $0.170 | |
Amazon Bedrockamazon-bedrock/eu-west-1 | No data | — | 99.97% | 98.34% | 99.10% | 469ms | 10.5s | 127 tps | 131K | $0.150 / $0.600 |
| No data | — | — | 70.95% | 85.99% | — | — | — | 131K | $0.200 / $0.950 | |
| No data | — | 91.27% | 86.45% | 88.19% | — | — | — | 131K | $0.150 / $0.600 |
- Amazon Bedrock100%◆ (best)Healthy24h 99.40%
- 7d
- 99.91%
- 30d
- 98.18%
- p50
- 603ms
- p99
- 17.0s
- Throughput
- 143 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
amazon-bedrock - BaseTen100%◆ (best)Healthy24h 99.97%
- 7d
- 99.85%
- 30d
- 99.93%
- p50
- 265ms
- p99
- 5.76s
- Throughput
- 124 tps
- Context
- 128K
- $/Mtok
- $0.100 / $0.500
baseten/fp4fp4 - Cerebras100%◆ (best)Healthy24h 99.98%
- 7d
- 99.95%
- 30d
- 99.95%
- p50
- 243ms
- p99
- 3.68s
- Throughput
- 745 tps◆ (best)
- Context
- 131K
- $/Mtok
- $0.350 / $0.750
cerebras/fp16fp16 - Crusoe100%◆ (best)Healthy24h 99.99%
- 7d
- 99.96%
- 30d
- 99.64%
- p50
- 208ms◆ (best)
- p99
- 5.45s
- Throughput
- 131 tps
- Context
- 131K
- $/Mtok
- $0.050 / $0.250
crusoe/bf16bf16 - DeepInfra100%◆ (best)Healthy24h 99.98%
- 7d
- 99.96%
- 30d
- 99.97%◆ (best)
- p50
- 561ms
- p99
- 2.05s
- Throughput
- 98 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
deepinfra/turbobf16 - DigitalOcean100%◆ (best)Healthy24h 99.98%
- 7d
- 99.94%
- 30d
- 99.94%
- p50
- 483ms
- p99
- 1.94s◆ (best)
- Throughput
- 25 tps
- Context
- 128K
- $/Mtok
- $0.060 / $0.420
digitalocean - Parasail100%◆ (best)Healthy24h 99.97%
- 7d
- 99.89%
- 30d
- 99.35%
- p50
- 406ms
- p99
- 1.98s
- Throughput
- 86 tps
- Context
- 131K
- $/Mtok
- $0.100 / $0.750
parasail/fp4fp4 - AkashML99.99%Healthy24h 99.95%
- 7d
- 99.97%◆ (best)
- 30d
- 99.91%
- p50
- 947ms
- p99
- 33.0s
- Throughput
- 32 tps
- Context
- 131K
- $/Mtok
- $0.037 / $0.187
akashml/bf16bf16 - DekaLLM99.93%Healthy24h 99.53%
- 7d
- 99.47%
- 30d
- 99.43%
- p50
- 901ms
- p99
- 49.8s
- Throughput
- 19 tps
- Context
- 131K
- $/Mtok
- $0.030 / $0.180◆ (best)
dekallm/bf16bf16 - Groq99.90%Healthy24h 99.40%
- 7d
- 99.08%
- 30d
- 99.64%
- p50
- 389ms
- p99
- 2.49s
- Throughput
- 274 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
groq - CoreWeave99.80%Healthy24h 98.92%
- 7d
- 98.99%
- 30d
- 99.68%
- p50
- 475ms
- p99
- 25.5s
- Throughput
- 32 tps
- Context
- 131K
- $/Mtok
- $0.030 / $0.170◆ (best)
coreweave/fp4fp4 - SambaNova99.71%Healthy24h 99.76%
- 7d
- 99.39%
- 30d
- 96.63%
- p50
- 800ms
- p99
- 7.58s
- Throughput
- 214 tps
- Context
- 131K
- $/Mtok
- $0.140 / $0.950
sambanova - Together98.10%Healthy24h 88.62%
- 7d
- 92.68%
- 30d
- 92.63%
- p50
- 261ms
- p99
- 4.95s
- Throughput
- 79 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
together - Mancer 293.27%Degraded24h 97.50%
- 7d
- 99.00%
- 30d
- 99.07%
- p50
- 1.50s
- p99
- 47.5s
- Throughput
- 25 tps
- Context
- 131K
- $/Mtok
- $0.045 / $0.250
mancer/fp8fp8 - Phala91.82%Degraded24h 98.96%
- 7d
- 99.31%
- 30d
- 92.31%
- p50
- 962ms
- p99
- 9.81s
- Throughput
- 93 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
phala - Novita90.75%Degraded24h 98.46%
- 7d
- 99.10%
- 30d
- 92.16%
- p50
- 845ms
- p99
- 19.4s
- Throughput
- 104 tps
- Context
- 131K
- $/Mtok
- $0.050 / $0.250
novita/fp4fp4 - Nebius88.95%Down24h 97.73%
- 7d
- 96.46%
- 30d
- 97.77%
- p50
- 441ms
- p99
- 11.5s
- Throughput
- 132 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
nebius/fp4fp4 - Google78.17%Down24h 66.73%
- 7d
- 54.88%
- 30d
- 78.16%
- p50
- 534ms
- p99
- 7.77s
- Throughput
- 101 tps
- Context
- 131K
- $/Mtok
- $0.090 / $0.360
google-vertex/global - Mara74.60%Down24h 95.80%
- 7d
- 93.43%
- 30d
- 85.85%
- p50
- 1.02s
- p99
- 10.9s
- Throughput
- 118 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.750
mara - DeepInfra64.61%Down24h 97.18%
- 7d
- 98.76%
- 30d
- 99.05%
- p50
- 13.1s
- p99
- 41.6s
- Throughput
- 13 tps
- Context
- 131K
- $/Mtok
- $0.037 / $0.170
deepinfra/bf16bf16 - No data24h 99.97%
- 7d
- 98.34%
- 30d
- 99.10%
- p50
- 469ms
- p99
- 10.5s
- Throughput
- 127 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
amazon-bedrock/eu-west-1 - No data24h —
- 7d
- 70.95%
- 30d
- 85.99%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 131K
- $/Mtok
- $0.200 / $0.950
deepinfra/fp8fp8 - No data24h 91.27%
- 7d
- 86.45%
- 30d
- 88.19%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
siliconflow/fp8fp8
Measured 1m 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.