Benchmark result

Ornith-1.5-9B-MTP-Q6_0_ROCMFPX_AGENT on AMD Radeon RX 6700/6700 XT/6750 XT / 6800M/6850M XT — 34.2 tok/s

Measured with LM Studio on August 24, 2026.

ShareRedditX

What the model built

Open full screen →

The coding scenario asks for a playable game in a single HTML file. This is exactly what the model returned, unedited; the frame only adds a line that tells this page how tall it is.

Reproduce this run

toolLM Studio
modelOrnith-1.5-9B-MTP-Q6_0_ROCMFPX_AGENT (Q6_0)
context8,192
clientv0.4.40+97

Set those in LM Studio, then:

llm-benchmark benchmark --model "Ornith-1.5-9B-MTP-Q6_0_ROCMFPX_AGENT" --tool "LM Studio"

The client prompts for context length and thinking mode, and for the KV cache dtype on Unsloth Studio. Sampling is left at the model default — the values above are what the tool reported using, not overrides the client sent.