Start typing to search.
Ranks providers by how quickly the first generated token appears. Providers without enough recent data remain visible after ranked rows.
NovitaAI currently ranks first at 1.49 s via novita/fp8.
| Rank | Provider / route | Time to first token | Speed measurements | Endpoint details |
|---|---|---|---|---|
| #1 | Provider / route NovitaAI novita/fp8 | Time to first token1.49 s | Speed measurements
| Endpoint details
|
| #2 | Provider / route SiliconFlow siliconflow/fp8 | Time to first token1.76 s | Speed measurements
| Endpoint details
|
| #3 | Provider / route AtlasCloud atlas-cloud/fp8 | Time to first token2.65 s | Speed measurements
| Endpoint details
|
Endpoint data fetched .
DeepSeek V3.2 Exp endpoint guide
3 of 3 DeepSeek V3.2 Exp endpoints currently have the published data required for this ranking. NovitaAI leads at 1.49 s via novita/fp8. The table keeps unranked routes visible so missing measurements do not look like missing provider availability.
The first-token ranking orders endpoints by the recent median time before output begins. This is useful for interactive applications, but it should be read alongside throughput because the first provider to start is not always the first to finish a long answer.
SiliconFlow currently ranks second at 1.76 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.
DeepSeek V3.2 Exp has 3 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.
