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.
Groq currently ranks first at 318.0 tok/s via groq.
| Rank | Provider / route | Output speed | Speed measurements | Endpoint details |
|---|---|---|---|---|
| #1 | Provider / route Groq groq | Output speed318.0 tok/s | Speed measurements
| Endpoint details
|
| #2 | Provider / route MARA mara | Output speed185.0 tok/s | Speed measurements
| Endpoint details
|
| #3 | Provider / route Fireworks fireworks | Output speed170.0 tok/s | Speed measurements
| Endpoint details
|
| #4 | Provider / route Together together/fp4 | Output speed156.0 tok/s | Speed measurements
| Endpoint details
|
| #5 | Provider / route SambaNova sambanova/minimax-m2.7-dedicated | Output speed112.0 tok/s | Speed measurements
| Endpoint details
|
| #6 | Provider / route SambaNova sambanova | Output speed103.5 tok/s | Speed measurements
| Endpoint details
|
| #7 | Provider / route DeepInfra deepinfra/turbo | Output speed58.5 tok/s | Speed measurements
| Endpoint details
|
| #8 | Provider / route MiniMax minimax/fp8 | Output speed53.0 tok/s | Speed measurements
| Endpoint details
|
| #9 | Provider / route GMICloud gmicloud/fp8 | Output speed49.0 tok/s | Speed measurements
| Endpoint details
|
| #10 | Provider / route MiniMax minimax/highspeed | Output speed39.5 tok/s | Speed measurements
| Endpoint details
|
| #11 | Provider / route NovitaAI novita/fp8 | Output speed38.0 tok/s | Speed measurements
| Endpoint details
|
| #12 | Provider / route DeepInfra deepinfra/fp8 | Output speed28.0 tok/s | Speed measurements
| Endpoint details
|
| #13 | Provider / route Morph morph | Output speed19.0 tok/s | Speed measurements
| Endpoint details
|
| #14 | Provider / route AtlasCloud atlas-cloud/fp8 | Output speed12.0 tok/s | Speed measurements
| Endpoint details
|
Endpoint data fetched .
MiniMax M2.7 endpoint guide
14 of 14 MiniMax M2.7 endpoints currently have the published data required for this ranking. Groq leads at 318.0 tok/s via groq. 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.
MARA currently ranks second at 185.0 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.
MiniMax M2.7 has 14 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.
