MobileLLM R1.5 950M — Hardware Requirements & GPU Compatibility
ChatReasoning
MobileLLM R1.5 950M is a 950M-parameter open language model from Meta. At BF16 it needs about 2.09 GB of VRAM — see which GPUs and Macs can run it below.
56 downloads
19 likesSpecifications
- Publisher
- Meta
- Parameters
- 950M
- Release Date
- 2025-11-24
- License
- Other
Get Started
HuggingFace
How Much VRAM Does MobileLLM R1.5 950M Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 2.1 GB | — | 1.90 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run MobileLLM R1.5 950M?
BF16 · 2.1 GBMobileLLM R1.5 950M (BF16) requires 2.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 3+ GB is recommended. 35 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti.
Runs great
— Plenty of headroomWhich Devices Can Run MobileLLM R1.5 950M?
BF16 · 2.1 GB33 devices with unified memory can run MobileLLM R1.5 950M, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Runs great
— Plenty of headroomRelated Models
Frequently Asked Questions
- How much VRAM does MobileLLM R1.5 950M need?
MobileLLM R1.5 950M requires 2.1 GB of VRAM at BF16.
VRAM = Weights + KV Cache + Overhead
Weights = 950M × 16 bits ÷ 8 = 1.9 GB
KV Cache + Overhead ≈ 0.2 GB (at 2K context + ~0.3 GB framework)
VRAM usage by quantization
BF162.1 GB- Can I run MobileLLM R1.5 950M on a Mac?
MobileLLM R1.5 950M requires at least 2.1 GB at BF16, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run MobileLLM R1.5 950M locally?
Yes — MobileLLM R1.5 950M can run locally on consumer hardware. At BF16 quantization it needs 2.1 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is MobileLLM R1.5 950M?
At BF16, MobileLLM R1.5 950M can reach ~1395 tok/s on AMD Instinct MI300X. On NVIDIA GeForce RTX 4090: ~314 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: AMD Instinct MI300X → 5300 ÷ 2.1 × 0.55 = ~1395 tok/s
Estimated speed at BF16 (2.1 GB)
~1395 tok/s~314 tok/s~1043 tok/s~862 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of MobileLLM R1.5 950M?
At BF16, the download is about 1.90 GB.
- Which GPUs can run MobileLLM R1.5 950M?
35 consumer GPUs can run MobileLLM R1.5 950M at BF16 (2.1 GB). Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 35 GPUs have plenty of headroom for comfortable inference.
- Which devices can run MobileLLM R1.5 950M?
33 devices with unified memory can run MobileLLM R1.5 950M at BF16 (2.1 GB), including Mac Mini M4 (16 GB), Mac Mini M4 (32 GB), Mac Mini M4 Pro (24 GB), Mac Mini M4 Pro (48 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.