Benchmark result
unsloth/LFM2.5-8B-A1B-GGUF on NVIDIA GeForce RTX 4070 — 259.7 tok/s
Measured with Unsloth Studio on September 28, 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
1st fastest of 3 runs of this model on NVIDIA GeForce RTX 4070 · median 241.2 tok/s.
sort
- 1259.7tok/s52.7Unsloth Studio · UD-Q5_K_XL · KV f16this run
- 2241.2tok/s52.6llama.cpp
- 3239.7tok/s44.8llama.cpp
unsloth/LFM2.5-8B-A1B-GGUF on other hardware
Reproduce this run
toolUnsloth Studio
modelunsloth/LFM2.5-8B-A1B-GGUF (UD-Q5_K_XL)
context100,000
kv cachef16
think limitnone
samplingtemperature 0.1 · top_p 0.1 · top_k 50 · min_p 0.15 · presence_penalty 0
clientv0.4.82+98
Set those in Unsloth Studio, then:
llm-benchmark benchmark --model "unsloth/LFM2.5-8B-A1B-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.