Contributors › Atlas-Inf's local-LLM benchmark configs
Atlas-Inf's local-LLM benchmark configs
Every tracked benchmark config measured by Atlas-Inf: model, quant, backend, tokens per second, source-linked and trust-tiered.
Snapshot 2026-10-15 · 394 configs on AI boxes and 423 on GPU cards · 123 models. Every number is generated from the tracker snapshot and linked to where it was measured; nothing here is typed by hand.
5 configs · 5 models · 28 stars · 131 tok/s best single-stream · Atlas-Inf on github.
Configs measured by Atlas-Inf
| Model | Hardware | Quant | Backend | Decode | Context | Trust | Source |
|---|---|---|---|---|---|---|---|
| Qwen3.5-35B-A3B | DGX Spark | NVFP4 | atlas (speculative decode, mtp-2) | 131 tok/s | Extracted, source-linked | source | |
| Qwen3.8-Flash-Next | DGX Spark | NVFP4 | atlas | 36.7 tok/s | Extracted, source-linked | source | |
| Qwen3.8-27B | DGX Spark | NVFP4 | atlas (speculative decode, mtp) | 23.6 tok/s | Extracted, source-linked | source | |
| Qwen3.5-27B | DGX Spark | NVFP4 | atlas | 13 tok/s | Extracted, source-linked | source |
More: all configs · hardware · news · contributors · methodology · llms.txt · llms-full.txt · API (OpenAPI) · JSON snapshot. Built by Altronis, private on-prem AI, Singapore.