Veyra 30M Base — Hardware Requirements & GPU Compatibility
Chat
Veyra 30M Base is a 35M-parameter open language model from veyra-ai. It supports a context window of up to 1,024 tokens. At Q4_K_M it needs about 0.33 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 · ~1986 tok/s)Macs from 8 GB (Q4_K_M · ~217 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Veyra 30M Base
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 veyra-ai/Veyra-30M-Base --max-model-len 1024vLLM repo with its own quantisation
Download: Weights on Hugging Face
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3080 Ti (12 GB) (11 GB headroom) · Q4_K_M
- Generation speed
- ~1797 tok/s
- generation speed
- Cost per 1M output tokens
- $0.019
- per 1M output tokens
How Much VRAM Does Veyra 30M Base 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 · ~2048 tok/s | |
| 0.3 GB | — | 3-bit medium — 3-bit medium quantization | Fits: — context · ~1986 tok/s | |
| 0.3 GB | — | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: — context · ~1986 tok/s | |
| 0.3 GB | — | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: — context · ~1986 tok/s | |
| 0.3 GB | — | 6-bit, very good — 6-bit quantization, very good quality | Fits: — context · ~1927 tok/s | |
| 0.3 GB | — | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: — context · ~1927 tok/s | |
| 0.4 GB | — | 16-bit · bfloat — Brain floating point 16 — preferred for training | Fits: — context · ~1724 tok/s |
Selected Q4_K_M
4-bit medium quantization0.3GBat 1K context
- Weights
- 0.1 GB
- KV cache
- 0.0 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 23.7 GB free · ~1986 tok/s
≥50 tok/s needs ≥25 GB/s · RTX 5090 ~3530, RTX 3090 Ti ~1986
Which Hardware Can Run Veyra 30M Base?
Q4_K_M · 0.3 GBVeyra 30M Base (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 Veyra 30M Base, 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 3080~1498Simulate output on RTX 3080RTX 2080 Ti~1213Simulate output on RTX 2080 TiRTX 3070 Ti~1198Simulate output on RTX 3070 TiM3*8 GB~217Simulate output on iMac 24" M3 (8 GB)Orin Nano~201Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~145Simulate output on iPhone 1720 GB
fits · 19.7 GB free · up to — ctx
1 GPU
RX 7900 XT~1455Simulate output on RX 7900 XT32–40 GB
fits · 31.7 GB free · up to — ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~3530Simulate output on RTX 5090M5 Max36 GB~976Simulate output on Mac Studio M5 Max (36 GB)M532 GB~326Simulate output on MacBook Air 13" M5 (32 GB)Orin~403Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~266Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Specifications
- Publisher
- veyra-ai
- Parameters
- 35M
- Architecture
- LlamaForCausalLM
- Context Length
- 1,024 tokens
- Vocabulary Size
- 8,192
- Release Date
- 2026-05-03
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does Veyra 30M Base need?
Veyra 30M Base requires 0.3 GB of VRAM at Q4_K_M, or 0.4 GB at BF16.
VRAM = Weights + KV Cache + Overhead
Weights = 35M × 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 Veyra 30M Base?
For Veyra 30M Base, 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 Veyra 30M Base on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run Veyra 30M Base. 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 Veyra 30M Base locally?
Yes — Veyra 30M Base can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Veyra 30M Base?
At Q4_K_M, Veyra 30M Base can reach ~14546 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~1986 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 ÷ 0.33 × 0.65 = ~15758 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 Veyra 30M Base?
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 0.01 GB.
Which GPUs can run Veyra 30M Base?
52 consumer GPUs can run Veyra 30M Base 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 Veyra 30M Base?
163 devices with unified memory can run Veyra 30M Base 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.