gpt-oss-120b
openai/gpt-oss-120b
Compare with gpt-oss-20b, GPT-5.4, GPT-5.5
| Provider | State | Uptime 5m | 24h | p50 latency | p99 | Throughput | Context | $/Mtok |
|---|---|---|---|---|---|---|---|---|
AkashMLakashml/bf16bf16 | Healthy | 100% | 99.15% | 555ms | 5.16s | 82 tps | 131K | $0.037 / $0.490 |
Amazon Bedrockamazon-bedrock | Healthy | 100% | 99.62% | 467ms | 1.63s | 179 tps | 131K | $0.150 / $0.600 |
Cerebrascerebras/fp16fp16 | Healthy | 100% | 99.99% | 244ms | 3.54s | 773 tps | 131K | $0.350 / $0.750 |
DigitalOceandigitalocean | Healthy | 100% | 99.83% | 881ms | 4.71s | 25 tps | 128K | $0.055 / $0.385 |
Mancer 2mancer/fp8fp8 | Healthy | 100% | 99.87% | 736ms | 6.23s | 56 tps | 131K | $0.085 / $0.500 |
Parasailparasail/fp4fp4 | Healthy | 100% | 99.42% | 398ms | 1.38s | 108 tps | 131K | $0.100 / $0.750 |
Phalaphala | Healthy | 100% | 99.89% | 847ms | 3.45s | 70 tps | 131K | $0.150 / $0.600 |
Groqgroq | Healthy | 99.91% | 99.90% | 263ms | 2.36s | 323 tps | 131K | $0.150 / $0.600 |
DeepInfradeepinfra/bf16bf16 | Healthy | 99.85% | 98.76% | 368ms | 13.8s | 47 tps | 131K | $0.037 / $0.170 |
Nebiusnebius/fp4fp4 | Degraded | 99.81% | 98.28% | 410ms | 11.3s | 158 tps | 131K | $0.150 / $0.600 |
CoreWeavecoreweave/fp4fp4 | Healthy | 99.65% | 96.05% | 480ms | 11.8s | 30 tps | 131K | $0.030 / $0.170 |
SambaNovasambanova | Healthy | 99.55% | 98.31% | 755ms | 4.59s | 386 tps | 131K | $0.140 / $0.950 |
DeepInfradeepinfra/turbobf16 | Healthy | 99.37% | 95.39% | 425ms | 1.44s | 146 tps | 131K | $0.150 / $0.600 |
Togethertogether | Healthy | 98.49% | 98.15% | 319ms | 5.11s | 34 tps | 131K | $0.150 / $0.600 |
Maramara | Healthy | 96.28% | 97.38% | 1.80s | 11.5s | 168 tps | 131K | $0.150 / $0.750 |
Novitanovita/fp4fp4 | Degraded | 94.53% | 96.67% | 606ms | 8.42s | 66 tps | 131K | $0.050 / $0.250 |
SiliconFlowsiliconflow/fp8fp8 | Degraded | 90.33% | 91.66% | 1.49s | 13.5s | 14 tps | 131K | $0.050 / $0.450 |
BaseTenbaseten/fp4fp4 | Down | 88.51% | 98.88% | 240ms | 1.79s | 181 tps | 128K | $0.100 / $0.500 |
Googlegoogle-vertex/global | Down | 86.94% | 90.20% | 1.11s | 5.21s | 56 tps | 131K | $0.090 / $0.360 |
Amazon Bedrockamazon-bedrock/eu-west-1 | No data | — | 100% | 346ms | 346ms | 122 tps | 131K | $0.150 / $0.600 |
- AkashML100%Healthy24h 99.15%
- p50
- 555ms
- p99
- 5.16s
- Throughput
- 82 tps
- Context
- 131K
- $/Mtok
- $0.037 / $0.490
akashml/bf16bf16 - Amazon Bedrock100%Healthy24h 99.62%
- p50
- 467ms
- p99
- 1.63s
- Throughput
- 179 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
amazon-bedrock - Cerebras100%Healthy24h 99.99%
- p50
- 244ms
- p99
- 3.54s
- Throughput
- 773 tps
- Context
- 131K
- $/Mtok
- $0.350 / $0.750
cerebras/fp16fp16 - DigitalOcean100%Healthy24h 99.83%
- p50
- 881ms
- p99
- 4.71s
- Throughput
- 25 tps
- Context
- 128K
- $/Mtok
- $0.055 / $0.385
digitalocean - Mancer 2100%Healthy24h 99.87%
- p50
- 736ms
- p99
- 6.23s
- Throughput
- 56 tps
- Context
- 131K
- $/Mtok
- $0.085 / $0.500
mancer/fp8fp8 - Parasail100%Healthy24h 99.42%
- p50
- 398ms
- p99
- 1.38s
- Throughput
- 108 tps
- Context
- 131K
- $/Mtok
- $0.100 / $0.750
parasail/fp4fp4 - Phala100%Healthy24h 99.89%
- p50
- 847ms
- p99
- 3.45s
- Throughput
- 70 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
phala - Groq99.91%Healthy24h 99.90%
- p50
- 263ms
- p99
- 2.36s
- Throughput
- 323 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
groq - DeepInfra99.85%Healthy24h 98.76%
- p50
- 368ms
- p99
- 13.8s
- Throughput
- 47 tps
- Context
- 131K
- $/Mtok
- $0.037 / $0.170
deepinfra/bf16bf16 - Nebius99.81%Degraded24h 98.28%
- p50
- 410ms
- p99
- 11.3s
- Throughput
- 158 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
nebius/fp4fp4 - CoreWeave99.65%Healthy24h 96.05%
- p50
- 480ms
- p99
- 11.8s
- Throughput
- 30 tps
- Context
- 131K
- $/Mtok
- $0.030 / $0.170
coreweave/fp4fp4 - SambaNova99.55%Healthy24h 98.31%
- p50
- 755ms
- p99
- 4.59s
- Throughput
- 386 tps
- Context
- 131K
- $/Mtok
- $0.140 / $0.950
sambanova - DeepInfra99.37%Healthy24h 95.39%
- p50
- 425ms
- p99
- 1.44s
- Throughput
- 146 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
deepinfra/turbobf16 - Together98.49%Healthy24h 98.15%
- p50
- 319ms
- p99
- 5.11s
- Throughput
- 34 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
together - Mara96.28%Healthy24h 97.38%
- p50
- 1.80s
- p99
- 11.5s
- Throughput
- 168 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.750
mara - Novita94.53%Degraded24h 96.67%
- p50
- 606ms
- p99
- 8.42s
- Throughput
- 66 tps
- Context
- 131K
- $/Mtok
- $0.050 / $0.250
novita/fp4fp4 - SiliconFlow90.33%Degraded24h 91.66%
- p50
- 1.49s
- p99
- 13.5s
- Throughput
- 14 tps
- Context
- 131K
- $/Mtok
- $0.050 / $0.450
siliconflow/fp8fp8 - BaseTen88.51%Down24h 98.88%
- p50
- 240ms
- p99
- 1.79s
- Throughput
- 181 tps
- Context
- 128K
- $/Mtok
- $0.100 / $0.500
baseten/fp4fp4 - Google86.94%Down24h 90.20%
- p50
- 1.11s
- p99
- 5.21s
- Throughput
- 56 tps
- Context
- 131K
- $/Mtok
- $0.090 / $0.360
google-vertex/global - Amazon Bedrock—No data24h 100%
- p50
- 346ms
- p99
- 346ms
- Throughput
- 122 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.600
amazon-bedrock/eu-west-1
Measured 2m ago. Latency and throughput are OpenRouter's rolling 30-minute windows.
History
1 endpoint reporting no data is omitted from the charts.
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 and p99 are OpenRouter's rolling 30-minute windows.
Throughput
Median tokens per second.