Benchmark result
unsloth/LFM2.5-1.2B-Instruct-GGUF on NVIDIA GeForce RTX 4070 — 333.4 tok/s
Measured with Unsloth Studio on September 26, 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 6 runs of this model on NVIDIA GeForce RTX 4070 · median 335.6 tok/s.
sort
- 1434.6tok/s38.7llama.cpp
- 2336.9tok/s32.7llama.cpp
- 3336.0tok/s30.1llama.cpp
- 4335.2tok/s31.6Unsloth Studio · Q4_K_M · KV q8_0
- 5333.4tok/s44.9Unsloth Studio · Q4_K_M · KV f16this run
- 6329.5tok/s35.2llama.cpp
unsloth/LFM2.5-1.2B-Instruct-GGUF on other hardware
Reproduce this run
toolUnsloth Studio
modelunsloth/LFM2.5-1.2B-Instruct-GGUF (Q4_K_M)
context65,536
kv cachef16
think limitnone
samplingtemperature 0.7 · top_p 0.95 · top_k -1 · min_p 0.01 · presence_penalty 0
clientv0.4.80+98
Set those in Unsloth Studio, then:
llm-benchmark benchmark --model "unsloth/LFM2.5-1.2B-Instruct-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.