AFM 4.5B — Hardware Requirements & GPU Compatibility
Chat
AFM 4.5B is a 4.6B-parameter open language model from Arcee AI. It supports a context window of up to 65,536 tokens. At Q4_K_M it needs about 3.22 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 · 64K)Best on 24 GB (Q4_K_M · ~204 tok/s · 64K)Macs from 8 GB (Q4_K_M · ~22 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run AFM 4.5B
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
vllm serve arcee-ai/AFM-4.5B --max-model-len 16384vLLM repo with its own quantisation
Download: Weights on Hugging Face
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3060 12GB (7 GB headroom) · Q4_K_M
- Generation speed
- ~73 tok/s
- generation speed
- Cost per 1M output tokens
- $0.23
- per 1M output tokens
How Much VRAM Does AFM 4.5B Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 2.4 GB | 4.7 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 64K context · ~272 tok/s | |
| 2.7 GB | 5 GB | 3-bit medium — 3-bit medium quantization | Fits: 64K context · ~243 tok/s | |
| 3.2 GB | 5.5 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 64K context · ~204 tok/s | |
| 3.7 GB | 6 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 64K context · ~175 tok/s | |
| 4.3 GB | 6.5 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 64K context · ~154 tok/s | |
| 5.1 GB | 7.3 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 64K context · ~129 tok/s | |
| 9.7 GB | 12 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Fits: 64K context · ~68 tok/s |
Selected Q4_K_M
4-bit medium quantization4.3GBat 16K context
- Weights
- 2.8 GB
- KV cache
- 1.2 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 19.7 GB free · ~204 tok/s
≥50 tok/s needs ≥247 GB/s · RTX 5090 ~362, RTX 3090 Ti ~204
Which Hardware Can Run AFM 4.5B?
Q4_K_M · 3.2 GBAFM 4.5B (Q4_K_M) requires 3.2 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 5+ GB is recommended. Using the full 66K context window can add up to 4.7 GB, bringing total usage to 7.9 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run AFM 4.5B, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), MacBook Air 13" M3 (8 GB).
≤ 11 GB
≤ 11 GB tier →fits · 3.7 GB free · up to 64K ctx
16 GPUs · 1 Mac class · 2 device families
RTX 3080~154Simulate output on RTX 3080RTX 2080 Ti~124Simulate output on RTX 2080 TiRTX 3070 Ti~123Simulate output on RTX 3070 TiM3*8 GB~22Simulate output on iMac 24" M3 (8 GB)Orin Nano~21Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~15Simulate output on iPhone 1720 GB
fits · 15.7 GB free · up to 64K ctx
1 GPU
RX 7900 XT~149Simulate output on RX 7900 XT32–40 GB
fits · 27.7 GB free · up to 64K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~362Simulate output on RTX 5090M5 Max36 GB~100Simulate output on Mac Studio M5 Max (36 GB)M532 GB~33Simulate output on MacBook Air 13" M5 (32 GB)Orin~41Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~27Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Qwen3.5 4BAlibaba | 98.5 | 3.4 GB | ~201 | same size |
| Gemma 3 4B ITGoogle | 64.0 | 5 GB | ~231 | same size |
| Translategemma 4B ITGoogle | — | 5 GB | ~231 | same size |
| Molmo2 4BAllen AI | — | 4.7 GB | ~193 | same size |
| Qwen3 VL 4B InstructAlibaba | — | 4.5 GB | ~208 | same size |
| Qwen3.6 27B MTPLX Optimized SpeedYoussofal | — | 4 GB | ~185 | same size |
Specifications
- Publisher
- Arcee AI
- Parameters
- 4.6B
- Architecture
- ArceeForCausalLM
- Context Length
- 65,536 tokens
- Vocabulary Size
- 128,005
- Release Date
- 2025-07-29
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does AFM 4.5B need?
AFM 4.5B requires 3.2 GB of VRAM at Q4_K_M, or 9.7 GB at BF16. Full 66K context adds up to 4.7 GB (7.9 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 4.6B × 4.8 bits ÷ 8 = 2.8 GB
KV Cache + Overhead ≈ 0.4 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 ≈ 5.1 GB (at full 66K context)
VRAM usage by quantization
What's the best quantization for AFM 4.5B?
For AFM 4.5B, Q4_K_M (3.2 GB) offers the best balance of quality and VRAM usage. Q5_K_M (3.7 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 2.4 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run AFM 4.5B on a Mac?
Yes — Mac Mini M2 (8 GB) and 145 other Macs can run AFM 4.5B. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 3.2 GB of usable unified memory (RAM minus macOS overhead).
Can I run AFM 4.5B locally?
Yes — AFM 4.5B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is AFM 4.5B?
At Q4_K_M, AFM 4.5B can reach ~1491 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~204 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 3.22 × 0.65 = ~1615 tok/s
Estimated speed at Q4_K_M (3.2 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of AFM 4.5B?
At Q4_K_M, the download is about 2.77 GB. The full-precision BF16 version is 9.24 GB. The smallest option (Q2_K) is 1.96 GB.
Which GPUs can run AFM 4.5B?
52 consumer GPUs can run AFM 4.5B 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 AFM 4.5B?
163 devices with unified memory can run AFM 4.5B at Q4_K_M (3.2 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.