Start typing to search.
Ranks providers by estimated time to return 500 tokens, combining the wait for the first token with output speed.
Cerebras currently ranks first at 0.82 s via cerebras/fp16.
| Rank | Provider / route | Estimated response time | Speed measurements | Endpoint details |
|---|---|---|---|---|
| #1 | Provider / route Cerebras cerebras/fp16 | Estimated response time0.82 s | Speed measurements
| Endpoint details
|
| #2 | Provider / route Groq groq | Estimated response time1.41 s | Speed measurements
| Endpoint details
|
| #3 | Provider / route Baseten baseten/fp4 | Estimated response time2.42 s | Speed measurements
| Endpoint details
|
| #4 | Provider / route Google Vertex google-vertex/global | Estimated response time3.81 s | Speed measurements
| Endpoint details
|
| #5 | Provider / route MARA mara | Estimated response time4.09 s | Speed measurements
| Endpoint details
|
| #6 | Provider / route SambaNova sambanova | Estimated response time4.17 s | Speed measurements
| Endpoint details
|
| #7 | Provider / route Together together | Estimated response time5.04 s | Speed measurements
| Endpoint details
|
| #8 | Provider / route Amazon Bedrock amazon-bedrock | Estimated response time5.32 s | Speed measurements
| Endpoint details
|
| #9 | Provider / route DeepInfra deepinfra/turbo | Estimated response time5.47 s | Speed measurements
| Endpoint details
|
| #10 | Provider / route Nebius nebius/fp4 | Estimated response time5.92 s | Speed measurements
| Endpoint details
|
| #11 | Provider / route Parasail parasail/fp4 | Estimated response time6.27 s | Speed measurements
| Endpoint details
|
| #12 | Provider / route Phala phala | Estimated response time7.95 s | Speed measurements
| Endpoint details
|
| #13 | Provider / route NovitaAI novita/fp4 | Estimated response time11.78 s | Speed measurements
| Endpoint details
|
| #14 | Provider / route OpenInference open-inference/int8 | Estimated response time13.60 s | Speed measurements
| Endpoint details
|
| #15 | Provider / route Weights & Biases wandb/fp4 | Estimated response time13.62 s | Speed measurements
| Endpoint details
|
| #16 | Provider / route Mancer mancer/fp4 | Estimated response time15.08 s | Speed measurements
| Endpoint details
|
| #17 | Provider / route DekaLLM dekallm/bf16 | Estimated response time16.74 s | Speed measurements
| Endpoint details
|
| #18 | Provider / route DeepInfra deepinfra/bf16 | Estimated response time20.95 s | Speed measurements
| Endpoint details
|
| #19 | Provider / route SiliconFlow siliconflow/fp8 | Estimated response time22.14 s | Speed measurements
| Endpoint details
|
| #20 | Provider / route DigitalOcean digitalocean | Estimated response time24.25 s | Speed measurements
| Endpoint details
|
| — | Provider / route Amazon Bedrock amazon-bedrock/eu-west-1 | Estimated response timeInsufficient recent data | Speed measurements
| Endpoint details
|
Endpoint data fetched .
gpt-oss-120b endpoint guide
20 of 21 gpt-oss-120b endpoints currently have the published data required for this ranking. Cerebras leads at 0.82 s via cerebras/fp16. The table keeps unranked routes visible so missing measurements do not look like missing provider availability.
The fastest ranking estimates a 500-token answer by adding the recent median first-token wait to 500 divided by median output throughput. It represents an example complete response, not a guarantee for every prompt or request size.
Groq currently ranks second at 1.41 s. Compare that gap with input and output price, response-start time, output speed, 30-minute uptime, context length, and the exact route before deciding whether first place is meaningful for your workload.
gpt-oss-120b has 21 published provider options, and performance data is matched to each exact OpenRouter routing tag. Different quantization, context, regional deployment, or provider configuration can change price and behavior even when the underlying model name is identical.
