LFM2.5 8B A1B DSpark — Hardware Requirements & GPU Compatibility
Chat
LFM2.5 8B A1B DSpark is a 8B-parameter open language model from Liquid AI in the LFM2.5 family. It supports a context window of up to 128,000 tokens. At Q4_K_M it needs about 5.12 GB of VRAM — see which GPUs and Macs can run it below.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M · 125K)Best on 24 GB (Q4_K_M · ~799 tok/s · 125K)Macs from 8 GB (Q3_K_M · ~120 tok/s)
Other
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run LFM2.5 8B A1B DSpark
One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.
Recommended
ollama run hf.co/LiquidAI/LFM2.5-8B-A1B-DSpark-GGUF:Q4_K_MQ4_K_M fits fully on the RTX 4090 · 18.7 GB free · 16K context
Download: LFM2.5-8B-A1B-DSpark-Q4_K_M.gguf0.20 GBLiquidAI
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3060 12GB (6 GB headroom) · Q4_K_M
- Generation speed
- ~332 tok/s
- generation speed
- Cost per 1M output tokens
- $0.051
- per 1M output tokens
How Much VRAM Does LFM2.5 8B A1B DSpark Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 3.7 GB | 4 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 125K context · ~1016 tok/s | |
| 4.2 GB | 4.5 GB | 3-bit medium — 3-bit medium quantization | Fits: 125K context · ~926 tok/s | |
| 5.1 GB | 5.4 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 125K context · ~799 tok/s | |
| 6.0 GB | 6.3 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 125K context · ~703 tok/s | |
| 6.9 GB | 7.2 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 125K context · ~627 tok/s | |
| 8.3 GB | 8.6 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 125K context · ~537 tok/s | |
| 16.3 GB | 16.6 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Fits: 125K context · ~295 tok/s |
Selected Q4_K_M
4-bit medium quantization5.3GBat 16K context
- Weights
- 4.8 GB
- KV cache
- 0.2 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 18.7 GB free · ~799 tok/s
≥50 tok/s needs ≥50 GB/s · RTX 5090 ~1213, RTX 3090 Ti ~799
Which Hardware Can Run LFM2.5 8B A1B DSpark?
Q4_K_M · 5.1 GBLFM2.5 8B A1B DSpark (Q4_K_M) requires 5.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 7+ GB is recommended. Using the full 128K context window can add up to 1.3 GB, bringing total usage to 6.4 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 151 devices with unified memory can run LFM2.5 8B A1B DSpark, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).
≤ 11 GB
≤ 11 GB tier →fits · 2.7 GB free · up to 125K ctx
16 GPUs · 1 device family
RTX 3080~637Simulate output on RTX 3080RTX 2080 Ti~534Simulate output on RTX 2080 TiRTX 3070 Ti~528Simulate output on RTX 3070 TiOrin Nano~101Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)20 GB
fits · 14.7 GB free · up to 125K ctx
1 GPU
RX 7900 XT~429Simulate output on RX 7900 XT32–40 GB
fits · 26.7 GB free · up to 125K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~1213Simulate output on RTX 5090M5 Max36 GB~335Simulate output on Mac Studio M5 Max (36 GB)M532 GB~144Simulate output on MacBook Air 13" M5 (32 GB)Orin~197Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~128Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Alternatives to LFM2.5 8B A1B DSpark
LFM2.5 family · 15 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| ALoDLM 8BAmazon | — | 7.6 GB | ~119 | same size |
| Llm Jp 4.1 8B Thinkingllm-jp | — | 7.6 GB | ~115 | same size |
| Ling 3.0 TinyInclusion AI | — | 5.5 GB | ~364 | same size |
| BitCPM CANN 8BOpenBMB | — | 5.6 GB | ~127 | same size |
| Gemma 4 E4BGoogle | — | 5.6 GB | ~123 | same size |
| InternVL3 5 8BOpenGVLab | — | 8.2 GB | ~116 | same size |
Specifications
- Publisher
- Liquid AI
- Family
- LFM2.5
- Parameters
- 8B
- Active parameters
- 1B per token
- Architecture
- Lfm2DSparkDraftModel
- Context Length
- 128,000 tokens
- Vocabulary Size
- 128,000
- Release Date
- 2026-08-10
- License
- Other
Frequently Asked Questions
How much VRAM does LFM2.5 8B A1B DSpark need?
LFM2.5 8B A1B DSpark requires 5.1 GB of VRAM at Q4_K_M, or 16.3 GB at BF16. Full 128K context adds up to 1.3 GB (6.4 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 8B × 4.8 bits ÷ 8 = 4.8 GB
KV Cache + Overhead ≈ 0.3 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
KV Cache + Overhead ≈ 1.6 GB (at full 128K context)
VRAM usage by quantization
What's the best quantization for LFM2.5 8B A1B DSpark?
For LFM2.5 8B A1B DSpark, Q4_K_M (5.1 GB) offers the best balance of quality and VRAM usage. Q5_K_M (6.0 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 3.7 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run LFM2.5 8B A1B DSpark on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run LFM2.5 8B A1B DSpark, the smallest at Q3_K_M. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 5.1 GB of usable unified memory (RAM minus macOS overhead), from Mac Mini M2 Pro (16 GB).
Can I run LFM2.5 8B A1B DSpark locally?
Yes — LFM2.5 8B A1B DSpark can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is LFM2.5 8B A1B DSpark?
At Q4_K_M, LFM2.5 8B A1B DSpark can reach ~882 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~799 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = 1000 ÷ (active GB ÷ (bandwidth GB/s × efficiency) × 1000 + layers × routing ms)
Mixture-of-Experts: just the active experts get read for each token, along with a fixed routing cost per layer.
Example: NVIDIA B200 → 0.6 GB active ÷ (8000 × 0.65) = 0.1231 ms, plus 5 layers × 0.055 ms = 0.2750 ms, so 1000 ÷ 0.3981 ms = ~2512 tok/s
Estimated speed at Q4_K_M (5.1 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of LFM2.5 8B A1B DSpark?
At Q4_K_M, the download is about 4.80 GB. The full-precision BF16 version is 16.00 GB. The smallest option (Q2_K) is 3.40 GB.
Which GPUs can run LFM2.5 8B A1B DSpark?
52 consumer GPUs can run LFM2.5 8B A1B DSpark at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 52 GPUs have plenty of headroom for comfortable inference.
Which devices can run LFM2.5 8B A1B DSpark?
151 devices with unified memory can run LFM2.5 8B A1B DSpark at Q4_K_M (5.1 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.