LLM Inference Benchmark Explorer
MiMo-V2.6-Flash-RL on 2× DGX Spark — FP4, vLLM, TP2, speculative decoding k=7 inference benchmark
Open this configuration in the LLM Inference Benchmark Explorer
| Model | MiMo-V2.6-Flash-RL |
|---|---|
| Parameters | 310B |
| Intelligence Index | — |
| Agentic Index | — |
| Device | 2× DGX Spark |
| Quantization | FP4 |
| Max C | 2 |
| Chat Capacity | 8 |
| Agentic Capacity | 3 |
| TP | 2 |
| DP | — |
| PP | — |
| Engine | vLLM |
| Speculative Decoding | Yes |
| C | TTFT (ms) | TPS (tok/s) | Status |
|---|---|---|---|
| 1 | 378 | 29.20 | PASS |
| 2 | 518 | 20.06 | PASS |
| 4 | 693 | 15.37 | FAIL |
| 8 | 876 | 10.61 | FAIL |
In OpenZeka's measurement, MiMo-V2.6-Flash-RL (310B parameters), served in FP4 format with vLLM and speculative decoding on 2× DGX Spark (TP=2), reached a generation speed of 29.2 tok/s per request and a time to first token (TTFT) of 378 ms with a single request.
Considering the speed targets and the available KV cache capacity, the estimated capacity is 8 users for chat use and 3 for agentic use. The realistic capacity will likely fall between these two values. In scenarios dominated by coding, tool use, long workflows and multi-agent use, capacity approaches the agentic estimate; where shorter interactions, standard conversations and lighter tasks dominate, it approaches the chat estimate.
Notes: DFlash speculative decoding k=7, 256K context. Served as MiMo-V2.6-Flash-RL. Model: XiaomiMiMo/MiMo-V2.6-Flash-RL
Intelligence Index and Agentic Index values are published by Artificial Analysis and are reproduced here with attribution.
In OpenZeka's measurement, MiMo-V2.6-Flash-RL (310B parameters), served in FP4 format with vLLM and speculative decoding on 2× DGX Spark (TP=2), reached a generation speed of 29.2 tok/s per request and a time to first token (TTFT) of 378 ms with a single request.
Considering the speed targets and the available KV cache capacity, the estimated capacity is 8 users for chat use and 3 for agentic use. The realistic capacity will likely fall between these two values. In scenarios dominated by coding, tool use, long workflows and multi-agent use, capacity approaches the agentic estimate; where shorter interactions, standard conversations and lighter tasks dominate, it approaches the chat estimate.