← CanIRunLocal · All hardware · All models
Yes, with partial offload (slower). LFM2.5 24B-A2B (MoE) on RTX 4060 Ti 8GB at Q4_K_M uses system RAM (16.2 GB needed, 7.5 GB usable), about 25 tok/s (estimate).
Hardware: RTX 4060 Ti 8GB · Model: LFM2.5 24B-A2B (MoE)
RTX 4060 Ti 8GB has 8 GB of VRAM and 288 GB/s of memory bandwidth. The checker holds back 0.5 GB of VRAM for the driver and assumes 16 GB of system RAM, so fast memory usable by Ollama is 7.5 GB. Partial offload may also use 60% of that system RAM. LFM2.5 24B-A2B (MoE) (24B, 2B active, Liquid AI) at Q4_K_M is a 15 GB download and needs 16.2 GB with headroom for about 8k tokens.
| Best quant | Q4_K_M |
|---|---|
| Download size | 15 GB |
| Memory needed | 16.2 GB |
| Usable fast memory | 7.5 GB |
| Fit | partial offload (slower) |
| Estimated speed | ~25 tok/s (estimate) |
| Command | ollama run lfm2.5:24b |
Estimates, not benchmarks. Model sizes are ollama.com download sizes; memory needed adds headroom for ~8k tokens of context; speeds are rough figures from memory bandwidth. Data updated 2026-10-23.
A project by OnDevice Labs · llms.txt