Start typing to search.
Ranks providers by their recent typical output speed in tokens per second. Providers without enough recent data remain visible after ranked rows.
Nebius currently ranks first at 36.0 tok/s via nebius/fp8.
| Rank | Provider / route | Output speed | Speed measurements | Endpoint details |
|---|---|---|---|---|
| #1 | Provider / route Nebius nebius/fp8 | Output speed36.0 tok/s | Speed measurements
| Endpoint details
|
| #2 | Provider / route Parasail parasail/fp8 | Output speed28.0 tok/s | Speed measurements
| Endpoint details
|
| #3 | Provider / route Friendli friendli | Output speed24.0 tok/s | Speed measurements
| Endpoint details
|
| #4 | Provider / route Google Vertex google-vertex/us-south1 | Output speed24.0 tok/s | Speed measurements
| Endpoint details
|
| #5 | Provider / route Google Vertex google-vertex/us-south1 | Output speed24.0 tok/s | Speed measurements
| Endpoint details
|
| #6 | Provider / route Alibaba Cloud Int. alibaba | Output speed22.0 tok/s | Speed measurements
| Endpoint details
|
| #7 | Provider / route DeepInfra deepinfra/fp8 | Output speed15.0 tok/s | Speed measurements
| Endpoint details
|
| #8 | Provider / route NovitaAI novita/fp8 | Output speed13.0 tok/s | Speed measurements
| Endpoint details
|
| #9 | Provider / route Venice venice/fp8 | Output speed8.0 tok/s | Speed measurements
| Endpoint details
|
| #10 | Provider / route StreamLake streamlake | Output speed8.0 tok/s | Speed measurements
| Endpoint details
|
| #11 | Provider / route GMICloud gmicloud/fp8 | Output speed2.0 tok/s | Speed measurements
| Endpoint details
|
| #12 | Provider / route AtlasCloud atlas-cloud/fp8 | Output speed0.0 tok/s | Speed measurements
| Endpoint details
|
Endpoint data fetched .
Qwen3 235B A22B Instruct 2507 endpoint guide
12 of 12 Qwen3 235B A22B Instruct 2507 endpoints currently have the published data required for this ranking. Nebius leads at 36.0 tok/s via nebius/fp8. The table keeps unranked routes visible so missing measurements do not look like missing provider availability.
The output-speed ranking compares recent median generated tokens per second. It favors routes that complete long generations quickly, while first-token latency remains a separate measure of how responsive the endpoint feels at the beginning.
Parasail currently ranks second at 28.0 tok/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.
Qwen3 235B A22B Instruct 2507 has 12 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.
