Phi 1 5 — Hardware Requirements & GPU Compatibility
ChatCode
Phi 1 5 is a 1.4B-parameter open language model from Microsoft in the Phi family. It supports a context window of up to 2,048 tokens. At Q4_K_M it needs about 0.94 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 · 2K)Best on 24 GB (Q4_K_M · ~697 tok/s · 2K)Macs from 8 GB (Q4_K_M · ~76 tok/s)
MIT
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Phi 1 5
One recommended way. Every other method, OS and quant is in the step-by-step guide.
Recommended
ollama run hf.co/mradermacher/phi-1_5-i1-GGUF:Q4_K_MQ4_K_M fits fully on the RTX 4090 · 23.1 GB free · 2K context
Download: phi-1_5.i1-Q4_K_M.gguf0.89 GBmradermacher
Other ways
each opens step-by-stepRun in cloud
Fits on RTX 3060 12GB (11 GB headroom) · Q4_K_M
- Generation speed
- ~249 tok/s
- generation speed
- Cost per 1M output tokens
- $0.093
- per 1M output tokens
How Much VRAM Does Phi 1 5 Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 0.7 GB | — | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 2K context · ~993 tok/s | |
| 0.7 GB | — | 3-bit small — 3-bit small quantization | Fits: 2K context · ~964 tok/s | |
| 0.8 GB | — | 3-bit medium — 3-bit medium quantization | Fits: 2K context · ~862 tok/s | |
| 0.8 GB | — | 4-bit · legacy — 4-bit legacy quantization | Fits: 2K context · ~840 tok/s | |
| 0.9 GB | — | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 2K context · ~697 tok/s | |
| 1.1 GB | — | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 2K context · ~590 tok/s | |
| 1.3 GB | — | 6-bit, very good — 6-bit quantization, very good quality | Fits: 2K context · ~508 tok/s | |
| 1.6 GB | — | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 2K context · ~420 tok/s |
Selected Q4_K_M
4-bit medium quantization0.9GBat 2K context
- Weights
- 0.4 GB
- KV cache
- 0.2 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 23.1 GB free · ~697 tok/s
≥50 tok/s needs ≥72 GB/s · RTX 5090 ~1239, RTX 3090 Ti ~697
Which Hardware Can Run Phi 1 5?
Q4_K_M · 0.9 GBPhi 1 5 (Q4_K_M) requires 0.9 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 2+ GB is recommended. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run Phi 1 5, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).
≤ 11 GB
≤ 11 GB tier →fits · 7.1 GB free · up to 2K ctx
16 GPUs · 1 Mac class · 2 device families
RTX 3080~526Simulate output on RTX 3080RTX 2080 Ti~426Simulate output on RTX 2080 TiRTX 3070 Ti~421Simulate output on RTX 3070 TiM3*8 GB~76Simulate output on iMac 24" M3 (8 GB)Orin Nano~71Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~51Simulate output on iPhone 1720 GB
fits · 19.1 GB free · up to 2K ctx
1 GPU
RX 7900 XT~511Simulate output on RX 7900 XT32–40 GB
fits · 31.1 GB free · up to 2K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~1239Simulate output on RTX 5090M5 Max36 GB~343Simulate output on Mac Studio M5 Max (36 GB)M532 GB~114Simulate output on MacBook Air 13" M5 (32 GB)Orin~142Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~93Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
How good is Phi 1 5?
llmrun Score blends public benchmark boards into one number; 100 = the average open model.
Not enough data yet for an llmrun Score.
A Score needs results on enough independent boards. How the Composite Score is built
Benchmarks
Here are all 6 boards that have results for this model, sorted by their weight in the llmrun Score; badges show who ran the test.
- MMLU
Epoch AI · Sep 2023Lab-reported
Lab-reported37.6%
#73/85 open
- HellaSwag
Epoch AI · Sep 2023Lab-reported
Lab-reported47.6%
#44/47 open
- ARC Challenge
Epoch AI · Sep 2023Lab-reported
Lab-reported44.4%
#37/56 open
- WinoGrande
Epoch AI · Sep 2023Lab-reported
Lab-reported73.4%
#19/51 open
- BoolQ
Epoch AI · Sep 2023Lab-reported
Lab-reported75.8%
#21/38 open
- OpenBookQA
Epoch AI · Sep 2023Lab-reported
Lab-reported37.2%
#22/26 open
Alternatives to Phi 1 5
Phi family · 9 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Inkling Small DSpark PreviewRadixArk | — | 1.5 GB | ~555 | same size |
| Sweep Next Edit 1.5BXenova | — | 1.6 GB | ~551 | same size |
| Ouro 1.4BByteDance | — | 6.4 GB | ~184 | same size |
| OLMo 2 0425 1B InstructAllen AI | — | 1.7 GB | ~449 | same size |
| Granite 3.0 1B A400m InstructIBM | — | 1.3 GB | ~524 | same size |
| Transformer 1.3B 100Bfla-hub | — | 3 GB | ~218 | same size |
Specifications
- Publisher
- Microsoft
- Family
- Phi
- Parameters
- 1.4B
- Architecture
- PhiForCausalLM
- Context Length
- 2,048 tokens
- Vocabulary Size
- 51,200
- Release Date
- 2023-09-10
- License
- MIT
Frequently Asked Questions
How much VRAM does Phi 1 5 need?
Phi 1 5 requires 0.9 GB of VRAM at Q4_K_M, or 3.1 GB at FP16.
VRAM = Weights + KV Cache + Overhead
Weights = 1.4B × 4.8 bits ÷ 8 = 0.9 GB
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
VRAM usage by quantization
What's the best quantization for Phi 1 5?
For Phi 1 5, Q4_K_M (0.9 GB) offers the best balance of quality and VRAM usage. Q5_K_S (1.1 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 0.4 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Phi 1 5 on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run Phi 1 5. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 0.9 GB of usable unified memory (RAM minus macOS overhead).
Can I run Phi 1 5 locally?
Yes — Phi 1 5 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Phi 1 5?
At Q4_K_M, Phi 1 5 can reach ~5106 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~697 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 ÷ 0.94 × 0.65 = ~5532 tok/s
Estimated speed at Q4_K_M (0.9 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Phi 1 5?
At Q4_K_M, the download is about 0.85 GB. The full-precision FP16 version is 2.84 GB. The smallest option (IQ2_XXS) is 0.39 GB.
Which GPUs can run Phi 1 5?
52 consumer GPUs can run Phi 1 5 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 Phi 1 5?
163 devices with unified memory can run Phi 1 5 at Q4_K_M (0.9 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.