Benchmark result
glm-4.7-flash-unsloth:small-hotfixed-90kHEAT on AMD Radeon RX 7900 XTX — 29.4 tok/s
Measured with Ollama 0.20.7 on April 17, 2026.
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.
How this run compares
5th fastest of 11 runs of this model on AMD Radeon RX 7900 XTX · median 28.6 tok/s.
sort
- 135.1tok/s60.6Ollama · Q3_K_S
- 234.7tok/s67.3Ollama · Q3_K_S
- 333.7tok/s70.1Ollama · Q3_K_S
- 430.5tok/s68.0Ollama · Q3_K_S
- 529.4tok/s65.2Ollama · Q3_K_Sthis run
- 628.6tok/s67.5Ollama · Q3_K_S
- 727.3tok/s75.5Ollama · Q3_K_S
- 825.5tok/s53.1Ollama · Q3_K_S
- 924.6tok/s68.9Ollama · Q3_K_S
- 1023.3tok/s66.9Ollama · Q3_K_S
- 116.9tok/s25.9Ollama · Q3_K_S
Reproduce this run
toolOllama 0.20.7
modelglm-4.7-flash-unsloth:small-hotfixed-90kHEAT (Q3_K_S)
context16,384
clientv0.4.12+88
Set those in Ollama, then:
llm-benchmark benchmark --model "glm-4.7-flash-unsloth:small-hotfixed-90kHEAT" --tool "Ollama"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.