Atomic Chat / LFM2.5 8B A1B needs about 7.3 GB of memory at Q4_K_M (a 5.2 GB file) with an 8K context window, so it fits on a 8 GB GPU. This repository has 12 quants, from 3.2 GB (Q2_K) to 9.0 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 6.4 GB.
Atomic Chat LFM2.5 8B A1B – a hybrid MoE model for on‑device assistants
Atomic Chat LFM2.5 8B A1B is a 8.3B‑parameter hybrid model with 1.5B active parameters per token. It supports a 128,000‑token context and is optimized for offline inference on llama.cpp, MLX, vLLM and SGLang. The model is multilingual and designed for agentic workflows and tool use.
Good at
- Agentic workflows and tool chaining
- Structured output generation
- Multilingual assistance
Limits
- No vision support
- Requires llama.cpp or compatible runtime
- License is other, not open source
Best for
- On‑device personal assistants
- Tool‑calling applications
- Complex instruction following
Who it’s for. Developers building offline, on‑device assistants that need reasoning and tool use. The model is suitable for applications requiring a large context window and multilingual support, but not for heavy programming or knowledge‑intensive QA without retrieval.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 3.2 GB | 5.3 GB | Fits | Get file → | |
| IQ3_M | 3.8 GB | 5.9 GB | Fits | Get file → | |
| Q3_K_M | 4.1 GB | 6.2 GB | Fits | Get file → | |
| Q3_K_L | 4.4 GB | 6.5 GB | Fits | Get file → | |
| IQ4_XS | 4.6 GB | 6.7 GB | Fits | Get file → | |
| Q4_K_S | 4.9 GB | 7.0 GB | Fits | Get file → | |
| Q4_K_M | 5.2 GB | 7.3 GB | Tight | Get file → | |
| Q4_K_XL | 5.2 GB | 7.3 GB | Tight | Get file → | |
| Q5_K_S | 5.9 GB | 8.0 GB | Tight | Get file → | |
| Q5_K_M | 6.0 GB | 8.1 GB | Too big | Get file → | |
| Q6_K | 7.0 GB | 9.1 GB | Too big | Get file → | |
| Q8_0 | 9.0 GB | 11.1 GB | Too big | Get file → |
For 8 GB VRAM, use Q4_K_M (5.16 GB file, ~7.3 GB RAM at 8 k context). For 12 GB VRAM, Q5_K_M (6.03 GB file, ~8.1 GB RAM). For maximum fidelity on 16 GB VRAM, Q6_K (6.96 GB file, ~9.1 GB RAM).
What GPU do you need for Atomic Chat / LFM2.5 8B A1B?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: the best quant that fits is IQ3_M (3.8 GB file, about 5.9 GB in use).
- 8 GB VRAM: the best quant that fits is Q5_K_S (5.9 GB file, about 8.0 GB in use).
- 12 GB VRAM: the best quant that fits is Q8_0 (9.0 GB file, about 11.1 GB in use).
Which Atomic Chat / LFM2.5 8B A1B quant should you choose?
Q4_K_M is the usual starting point for Atomic Chat / LFM2.5 8B A1B: it balances file size and quality well. Here it is a 5.2 GB file.
The Q8_0 file is 9.0 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q2_K at 3.2 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Atomic Chat / LFM2.5 8B A1B
How much VRAM does Atomic Chat / LFM2.5 8B A1B need?
About 7.3 GB at Q4_K_M with an 8K context window. That is the 5.2 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Atomic Chat / LFM2.5 8B A1B on an 8 GB GPU?
Yes, with Q5_K_S: the 5.9 GB file needs about 8.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Atomic Chat / LFM2.5 8B A1B GGUF?
It depends on the quant: Q2_K is 3.2 GB, IQ3_M is 3.8 GB, Q3_K_M is 4.1 GB, Q3_K_L is 4.4 GB, IQ4_XS is 4.6 GB, Q4_K_S is 4.9 GB and 6 more.
What context length does Atomic Chat / LFM2.5 8B A1B support?
The GGUF header lists a trained context length of 128,000 tokens (125K). The KV cache grows with the context you use: about 1.6 GB at 8K and 6.4 GB at 32K.
Where can I download Atomic Chat / LFM2.5 8B A1B GGUF?
The files are on Hugging Face in the AlexAtomic/lfm25-8b-a1b-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Atomic Chat / LFM2.5 8B A1B use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run LFM2.5 8B A1B on 8 GB VRAM?
Yes, using the Q4_K_M quantization which requires about 7.3 GB of RAM at 8 k context.
Does this model support vision tasks?
No, it is a text‑only model.
What runtime is required for offline inference?
It can be run with llama.cpp, MLX, vLLM or SGLang with the --jinja flag for the chat template.