Stories15M MOE — Hardware Requirements & GPU Compatibility
Chat
Stories15M MOE is a 36M-parameter open language model from ggml-org. It supports a context window of up to 256 tokens. At Q4_K_M it needs about 0.34 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)Best on 24 GB (Q4_K_M · ~1360 tok/s)Macs from 8 GB (Q4_K_M · ~204 tok/s)
MIT
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Stories15M MOE
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 ggml-org/stories15M_MOE --max-model-len 256vLLM 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 (11 GB headroom) · Q4_K_M
- Generation speed
- ~683 tok/s
- generation speed
- Cost per 1M output tokens
- $0.025
- per 1M output tokens
How Much VRAM Does Stories15M MOE Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 0.3 GB | — | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: — context · ~1382 tok/s | |
| 0.3 GB | — | 3-bit medium — 3-bit medium quantization | Fits: — context · ~1382 tok/s | |
| 0.3 GB | — | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: — context · ~1360 tok/s | |
| 0.3 GB | — | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: — context · ~1360 tok/s | |
| 0.3 GB | — | 6-bit, very good — 6-bit quantization, very good quality | Fits: — context · ~1360 tok/s | |
| 0.3 GB | — | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: — context · ~1338 tok/s | |
| 0.4 GB | — | 16-bit · bfloat — Brain floating point 16 — preferred for training | Fits: — context · ~1258 tok/s |
Selected Q4_K_M
4-bit medium quantization0.3GBat 0K context
- Weights
- 0.1 GB
- KV cache
- 0.0 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 23.7 GB free · ~1360 tok/s
≥50 tok/s needs ≥21 GB/s · RTX 5090 ~1792, RTX 3090 Ti ~1360
Which Hardware Can Run Stories15M MOE?
Q4_K_M · 0.3 GBStories15M MOE (Q4_K_M) requires 0.3 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 1+ 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 Stories15M MOE, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).
≤ 11 GB
≤ 11 GB tier →fits · 7.7 GB free · up to — ctx
16 GPUs · 1 Mac class · 2 device families
RTX 3080~1153Simulate output on RTX 3080RTX 2080 Ti~1007Simulate output on RTX 2080 TiRTX 3070 Ti~998Simulate output on RTX 3070 TiM3*8 GB~204Simulate output on iMac 24" M3 (8 GB)Orin Nano~231Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~148Simulate output on iPhone 1720 GB
fits · 19.7 GB free · up to — ctx
1 GPU
RX 7900 XT~570Simulate output on RX 7900 XT32–40 GB
fits · 31.7 GB free · up to — ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~1792Simulate output on RTX 5090M5 Max36 GB~494Simulate output on Mac Studio M5 Max (36 GB)M532 GB~273Simulate output on MacBook Air 13" M5 (32 GB)Orin~430Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~276Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Specifications
- Publisher
- ggml-org
- Parameters
- 36M
- Active parameters
- 0B per token
- Architecture
- MixtralForCausalLM
- Context Length
- 256 tokens
- Vocabulary Size
- 32,000
- Release Date
- 2024-07-11
- License
- MIT
Frequently Asked Questions
How much VRAM does Stories15M MOE need?
Stories15M MOE requires 0.3 GB of VRAM at Q4_K_M, or 0.4 GB at BF16.
VRAM = Weights + KV Cache + Overhead
Weights = 36M × 4.8 bits ÷ 8 = 0 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.
VRAM usage by quantization
What's the best quantization for Stories15M MOE?
For Stories15M MOE, Q4_K_M (0.3 GB) offers the best balance of quality and VRAM usage. Q5_K_M (0.3 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 0.3 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Stories15M MOE on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run Stories15M MOE. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 0.3 GB of usable unified memory (RAM minus macOS overhead).
Can I run Stories15M MOE locally?
Yes — Stories15M MOE can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Stories15M MOE?
At Q4_K_M, Stories15M MOE can reach ~797 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~1360 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 → 0.3 GB active ÷ (8000 × 0.65) = 0.0511 ms, plus 6 layers × 0.055 ms = 0.3300 ms, so 1000 ÷ 0.3811 ms = ~2624 tok/s
Estimated speed at Q4_K_M (0.3 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Stories15M MOE?
At Q4_K_M, the download is about 0.02 GB. The full-precision BF16 version is 0.07 GB. The smallest option (Q2_K) is 0.02 GB.
Which GPUs can run Stories15M MOE?
52 consumer GPUs can run Stories15M MOE 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 Stories15M MOE?
163 devices with unified memory can run Stories15M MOE at Q4_K_M (0.3 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.