LFM2.5-8B-A1B-Abliterated
on a Mac: what it actually needs

For PocketAIHub. Worked out from your repo, no sign-up, nothing to click.

4-bit
24 layers
8 KV heads
head dim 64
2 likes
495 downloads
5.5 GB
measureddownload size, summed from the 2 safetensors files in your repo
16 GB Mac
computedthe smallest machine that holds it at 8k context, assuming MLX can wire 75% of unified memory

What it needs as context grows

4k context0.20 GB KV + 5.5 GB weights + 1.2 GB runtime = 6.9 GB, smallest Mac 16 GB
8k context0.40 GB KV + 5.5 GB weights + 1.2 GB runtime = 7.1 GB, smallest Mac 16 GB
32k context1.61 GB KV + 5.5 GB weights + 1.2 GB runtime = 8.3 GB, smallest Mac 16 GB

KV cache = 2 tensors x 24 layers x 8 KV heads x 64 head dim x 2 bytes (fp16) x tokens. Read straight off your config.json. Runtime overhead is the 1.2 GB I see for Python, MLX, the tokenizer and activations on my own runs.

Mac by Mac

ContextNeeds81618243236486496128
4k6.9 GBnoyesyesyesyesyesyesyesyesyes
8k7.1 GBnoyesyesyesyesyesyesyesyesyes
32k8.3 GBnoyesyesyesyesyesyesyesyesyes

Columns are unified memory in GB. "Tight" means it fits but leaves under 2 GB for anything else, so expect swap once a browser is open.

Run it

pip install mlx-lm
mlx_lm.generate --model PocketAiHub/LFM2.5-8B-A1B-Abliterated \
  --prompt "hello" --max-tokens 100

What is not on this page

not measuredTokens per second. That needs a real run on real silicon and I will not print a number I have not seen. If you want it, reply and I will run LFM2.5-8B-A1B-Abliterated on an M5 and add the result here with the command I used.
Speed, measured: this page is about memory. For how fast models actually run on Apple Silicon, with the tasks and judging open, see the Agent-12 local agent leaderboard.

Built by Matt Macosko, who publishes 29 MLX quants and wrote claude-code-local, which runs Claude Code fully on-device on Apple Silicon. Source for this page: your model repo and its config.