Benchmark result
unsloth/LFM2.5-1.2B-Thinking-GGUF on NVIDIA GeForce RTX 4070 — 335.4 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
1st fastest of 8 runs of this model on NVIDIA GeForce RTX 4070 · median 323.3 tok/s.
sort
- 1335.4tok/s31.8Unsloth Studio · Q4_K_M · KV q8_0this run
- 2332.8tok/s28.2llama.cpp
- 3329.7tok/s32.3llama.cpp
- 4328.2tok/s38.9llama.cpp
- 5318.5tok/s31.2llama.cpp
- 6314.3tok/s30.2llama.cpp
- 7309.0tok/s29.1llama.cpp
- 8304.0tok/s27.8llama.cpp
unsloth/LFM2.5-1.2B-Thinking-GGUF on other hardware
Reproduce this run
toolUnsloth Studio
modelunsloth/LFM2.5-1.2B-Thinking-GGUF (Q4_K_M)
context65,536
kv cacheq8_0
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-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.