How fast is prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0 for local inference?

Across 3 community benchmark runs, prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0 reaches up to 225.9 tok/s and averages 222.0 tok/s, with the fastest results on NVIDIA GeForce RTX 4070.

How much memory does prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0 need?

The leanest observed configuration used about 7.7 GB of memory (quantizations tested: Q2_0).

Which tools have been used to run prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0?

Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.

Model benchmarks

prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0 local LLM performance

As of October 2026, prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0 runs at up to 225.9 tok/s for local inference (best of 3 community benchmark runs across 1 GPU).

llama.cppQ2_0

ShareReddit X

Model size

Peak speed

225.9 tok/s

Average speed

222.0 tok/s

Avg PP

4320.2 tok/s

Min memory

7.7 GB

Max context

262,144 tokens

Avg output / run

16,801 tokens

Avg runtime / run

1m 19s

Avg quality

52.7

Benchmark runs

GPUs tested

Quality by task

Average LLM-judged quality (0–100) with the run-to-run spread shown as a P5–P95 band, overall and for each benchmark task, across all 3 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.

Task	P5 (low)	Avg	P95 (high)
Overall	46.3	52.7	59.3
Agent Workflow	44.9	61.3	71.0
Code Generation	9.7	20.0	29.0
Role Play & Narrative	63.7	72.4	78.8
Research & Analysis	47.8	57.1	66.9

Performance by hardware and tool

Every hardware/tool/quantization combination prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0 has been benchmarked on, ranked by peak token generation speed. Last updated October 2026.

Hardware	Tool	Quant	Best tok/s	Avg tok/s	Memory	Context	Quality	Runs
NVIDIA GeForce RTX 4070	llama.cpp	Q2_0	225.9 tok/s	222.0 tok/s	7.7 GB	262,144 tokens	52.7	3

Benchmark runs

All 3 prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0 runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.

Open run page →

Frequently asked questions

How fast is prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0 for local inference?: Across 3 community benchmark runs, prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0 reaches up to 225.9 tok/s and averages 222.0 tok/s, with the fastest results on NVIDIA GeForce RTX 4070.
How much memory does prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0 need?: The leanest observed configuration used about 7.7 GB of memory (quantizations tested: Q2_0).
Which tools have been used to run prism-ml/Ternary-Bonsai-2-4B-gguf:Q2_0?: Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.