MedErtas v0.1 31B LS — Hardware Requirements & GPU Compatibility
ChatFunctions
MedErtas v0.1 31B LS is a 20.9B-parameter open language model from ErtasAI. It supports a context window of up to 262,144 tokens. At Q4_K_M it needs about 14.14 GB of VRAM — see which GPUs and Macs can run it below.
Runs well on a 24 GB GPU
Runs from 12 GB (Q2_K · 16K)Best on 24 GB (Q4_K_M · ~46 tok/s · 64K)Macs from 16 GB (Q2_K · ~13 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run MedErtas v0.1 31B LS
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 ErtasAI/MedErtas-v0.1-31B-LS --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 3090 (24 GB) (8 GB headroom) · Q4_K_M
- Generation speed
- ~43 tok/s
- generation speed
- Cost per 1M output tokens
- $1.19
- per 1M output tokens
How Much VRAM Does MedErtas v0.1 31B LS Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 10.5 GB | 13.9 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 64K context · ~63 tok/s | |
| 11.8 GB | 15.2 GB | 3-bit medium — 3-bit medium quantization | Fits: 64K context · ~56 tok/s | |
| 14.1 GB | 17.6 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 64K context · ~46 tok/s | |
| 16.5 GB | 19.9 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 32K context · ~40 tok/s | |
| 18.8 GB | 22.3 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 32K context · ~35 tok/s | |
| 22.5 GB | 25.9 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 16K context · ~29 tok/s | |
| 43.4 GB | 46.8 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Does not fit: needs 44.7 GB |
Selected Q4_K_M
4-bit medium quantization15.5GBat 16K context
- Weights
- 12.5 GB
- KV cache
- 2.6 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 8.5 GB free · ~46 tok/s
≥50 tok/s needs ≥1.09 TB/s · RTX 5090 ~82
Which Hardware Can Run MedErtas v0.1 31B LS?
Q4_K_M · 14.1 GBMedErtas v0.1 31B LS (Q4_K_M) requires 14.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 19+ GB is recommended. Using the full 262K context window can add up to 33.0 GB, bringing total usage to 47.2 GB. 26 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 5080. 124 devices with unified memory can run MedErtas v0.1 31B LS, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Mini M5 Pro (24 GB).
16–18 GB
16–18 GB tier →fits · 0.5 GB free · up to 16K ctx
18 GPUs · 1 Mac class
RTX 5080~44Simulate output on RTX 5080RTX 5070 Ti~41Simulate output on RTX 5070 TiRTX 4080 SUPER~34Simulate output on RTX 4080 SUPERM3 Pro*18 GB~7Simulate output on MacBook Pro 14" M3 Pro (18 GB)20 GB
fits · 4.5 GB free · up to 32K ctx
1 GPU
RX 7900 XT~34Simulate output on RX 7900 XT32–40 GB
fits · 16.5 GB free · up to 128K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~82Simulate output on RTX 5090M5 Max36 GB~23Simulate output on Mac Studio M5 Max (36 GB)M532 GB~8Simulate output on MacBook Air 13" M5 (32 GB)Orin~9Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~6Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Similar models
4| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| InternVL3 5 GPT OSS 20B A4B Preview HFOpenGVLab | — | 13.6 GB | ~180 | same size |
| Internlm 20BInternLM | — | 45.3 GB | —needs > 24 GB | same size |
| ERNIE 4.5 21B A3B ThinkingBaidu | — | 14.3 GB | ~213 | same size |
| Mistral Small Instruct 2409Mistral AI | — | 17.4 GB | ~46 | same size |
Specifications
- Publisher
- ErtasAI
- Parameters
- 20.9B
- Architecture
- Gemma4ForConditionalGeneration
- Context Length
- 262,144 tokens
- Vocabulary Size
- 262,144
- Release Date
- 2026-11-09
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does MedErtas v0.1 31B LS need?
MedErtas v0.1 31B LS requires 14.1 GB of VRAM at Q4_K_M, or 43.4 GB at BF16. Full 262K context adds up to 33.0 GB (47.2 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 20.9B × 4.8 bits ÷ 8 = 12.5 GB
KV Cache + Overhead ≈ 1.6 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 ≈ 34.7 GB (at full 262K context)
VRAM usage by quantization
Can NVIDIA GeForce RTX 4090 run MedErtas v0.1 31B LS?
Yes, at Q8_0 (22.5 GB) or lower. Higher quantizations like BF16 (43.4 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.
What's the best quantization for MedErtas v0.1 31B LS?
For MedErtas v0.1 31B LS, Q4_K_M (14.1 GB) offers the best balance of quality and VRAM usage. Q5_K_M (16.5 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 10.5 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run MedErtas v0.1 31B LS on a Mac?
Yes — Mac Mini M2 Pro (16 GB) and 133 other Macs can run MedErtas v0.1 31B LS, the smallest at Q2_K. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 14.1 GB of usable unified memory (RAM minus macOS overhead), from MacBook Pro 14" M3 Pro (18 GB).
Can I run MedErtas v0.1 31B LS locally?
Yes — MedErtas v0.1 31B LS can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is MedErtas v0.1 31B LS?
At Q4_K_M, MedErtas v0.1 31B LS can reach ~340 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~46 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 ÷ 14.14 × 0.65 = ~368 tok/s
Estimated speed at Q4_K_M (14.1 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of MedErtas v0.1 31B LS?
At Q4_K_M, the download is about 12.52 GB. The full-precision BF16 version is 41.74 GB. The smallest option (Q2_K) is 8.87 GB.
Which GPUs can run MedErtas v0.1 31B LS?
26 consumer GPUs can run MedErtas v0.1 31B LS at Q4_K_M. Top options include AMD Radeon RX 7900 XTX, NVIDIA GeForce RTX 3090, NVIDIA GeForce RTX 3090 Ti, AMD Radeon RX 6800. 7 GPUs have plenty of headroom for comfortable inference.
Which devices can run MedErtas v0.1 31B LS?
124 devices with unified memory can run MedErtas v0.1 31B LS at Q4_K_M (14.1 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.