Benchmark result
JetBrains/Mellum2.1-12B-A2.5B-Thinking-GGUF on NVIDIA GeForce RTX 4070 — 37.7 tok/s
Measured with Unsloth Studio on October 12, 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 5 runs of this model on NVIDIA GeForce RTX 4070 · median 151.5 tok/s.
sort
- 1173.3tok/s66.0Unsloth Studio · Q4_K_M · KV q4_0
- 2167.4tok/s69.0Unsloth Studio · Q4_K_M · KV q8_0
- 3151.5tok/s63.4Unsloth Studio · Q4_K_M · KV q5_0
- 444.8tok/s65.4Unsloth Studio · Q4_K_M · KV q4_1
- 537.7tok/s62.3Unsloth Studio · Q4_K_M · KV q5_1this run
Reproduce this run
toolUnsloth Studio
modelJetBrains/Mellum2.1-12B-A2.5B-Thinking-GGUF (Q4_K_M)
context65,536
kv cacheq5_1
think limitnone
samplingtemperature 0.7 · top_p 0.95 · top_k -1 · min_p 0.01 · presence_penalty 0
clientv0.4.92+102
Set those in Unsloth Studio, then:
llm-benchmark benchmark --model "JetBrains/Mellum2.1-12B-A2.5B-Thinking-GGUF" --tool "Unsloth 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.