by tonyd2wild · Agent Tool · ★ 28
MiMo-V2.5 Omni · TP=2 · 1M context · NVFP4 KV on 2× DGX Spark Running (Omni: text + image + video + audio) tensor-parallel across two NVIDIA DGX Spark (GB10) boxes, with 4-bit KV cache + MTP speculative decoding — serving a 1,000,000-token context with a 1.97M-token KV pool. This is the 2-node sibling of the 3-node build (MiMo-V2.5-TP3-NVFP4-KV-3xDGX-Spark). Two Sparks instead of three — so it pairs cleanly with another 2-node model on the same fleet (e.g. DeepSeek-V4 TP=2 on the other pair).
| Stars | 28 |
| Forks | 3 |
| Language | Python |
| Category | Agent Tool |
| Quality Score | 63.1333546791928/100 |
| Last Updated | 2026-07-01 |
| Created | 2026-06-23 |
| Platforms | python |
| Est. Tokens | ~17k |
Explore other popular agent tool tools:
MiMo-V2.5-TP2-1M-NVFP4-KV-2xDGX-Spark is MiMo-V2.5 Omni TP=2 on 2x DGX Spark · 1M context · NVFP4 4-bit KV (~1.97M-token KV pool @ 1M, ~30 tok/s) · 69-eval: thinking-OFF 97.8 beats thinking-ON 90.6 for tool/agent work. It is categorized as a Agent Tool with 28 GitHub stars.
MiMo-V2.5-TP2-1M-NVFP4-KV-2xDGX-Spark is primarily written in Python.
You can find installation instructions and usage details in the MiMo-V2.5-TP2-1M-NVFP4-KV-2xDGX-Spark GitHub repository at github.com/tonyd2wild/MiMo-V2.5-TP2-1M-NVFP4-KV-2xDGX-Spark. The project has 28 stars and 3 forks, indicating an active community.