llamaperf

Bonsai 2 27B

on NVIDIA RTX 4060 Laptop 8GB · Kindi · 24,576 ctx

NVIDIA hardware
Oct 31, 2026
Throughput
28.0 t/s gen · 31.2 tokens/s prompt processing (prefill)Prompt processing measures input prefill speed; it does not tell you the time to first token.
Quant
PTQ1_0 (GGUF)
VRAM reported
8 GB

Summary

User reports Bonsai 2 27B at 28.0 tok/s on an RTX 4060 Laptop 8GB, with prompt reading at 31.2 tokens/s. Setup is the Kindi 0.3.0 image with a PTQ1_0 GGUF at 24,576 context, leaving 1,367 MiB free. Decode falls from 28.0 tok/s at position 100 to 25.2 tok/s at 24,000; a verify run passed token-exact against llama.cpp at 29.3 tokens/s.