llmrun
ConwayResearch

Underdog Saluki 27B 1.0 — Hardware Requirements & GPU Compatibility

ChatFunctions

Underdog Saluki 27B 1.0 is a 27B-parameter open language model from ConwayResearch. At Q4_K_M it needs about 17.82 GB of VRAM — see which GPUs and Macs can run it below.

27B params · denseApache 2.0269.8K downloads400 likes
Based on Qwen3.8 27B
B

Decent on a 24 GB GPU

Runs on24 GB Q4_K_M · ~37 tok/s · 16KMacs from32 GB ~5 tok/s

Apache 2.0

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

Run Underdog Saluki 27B 1.0

One recommended way. Every other method, OS and quant is in the step-by-step guide.

Recommended

vLLM
one command · needs vLLM installed
$vllm serve ConwayResearch/Underdog-Saluki-27B-1.0 --max-model-len 16384

vLLM repo with its own quantisation

Download: Weights on Hugging Face

install, OS, other quants

Other ways

each opens step-by-step

Run in cloud

Fits on RTX A6000 (48 GB) (26 GB headroom) · Q4_K_M

Generation speed
~28 tok/s
generation speed
Cost per 1M output tokens
$3.27
per 1M output tokens
Deployon RunPodBest optionfrom $0.33/hr (opens in new tab)
Compare GPUs →
or

How Much VRAM Does Underdog Saluki 27B 1.0 Need?

Select a quantization level; the listed hardware adheres to it.

QuantizationVRAM+32KOn RTX 4090 (reference)
12.6 GB20.4 GB2-bit · K-quant — 2-bit quantization with K-quant improvementsFits: 32K context · ~52 tok/s
14.5 GB22.2 GB3-bit medium — 3-bit medium quantizationFits: 32K context · ~45 tok/s
17.8 GB25.6 GB4-bit medium — 4-bit medium quantization — most popular sweet spotFits: 16K context · ~37 tok/s
21.2 GB28.9 GB5-bit medium — 5-bit medium quantization — good quality/size tradeoffFits: 8K context · ~31 tok/s
24.5 GB32.2 GB6-bit, very good — 6-bit quantization, very good qualityDoes not fit: needs 28.1 GB
29.7 GB37.4 GB8-bit, near-lossless — 8-bit quantization, near-losslessDoes not fit: needs 33.3 GB
59.4 GB67.1 GB16-bit · bfloat — Brain floating point 16 — preferred for trainingDoes not fit: needs 63 GB
est.calculated, no published file yet

Selected Q4_K_M

4-bit medium quantization

21.4GBat 16K context

Weights
17.0 GB
KV cache
4.1 GB
Overhead
0.3 GB

Fits RTX 4090 (reference) · 2.6 GB free · ~37 tok/s

≥50 tok/s needs ≥1.37 TB/s · RTX 5090 ~65

Which Hardware Can Run Underdog Saluki 27B 1.0?

Q4_K_M · 17.8 GB

Underdog Saluki 27B 1.0 (Q4_K_M) requires 17.8 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 24+ GB is recommended. 8 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 125 devices with unified memory can run Underdog Saluki 27B 1.0, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Mini M5 Pro (24 GB).

By VRAM tier · typical GPUs and Macs at Q4_K_M
8–16 GB · RTX 4060, RTX 3060 12GB, RTX 5060 Ti 16GBDoesn't fit — smallest quant needs 16.2 GB8 GB tier →12 GB tier →16 GB tier →
24 GB · RTX 4090referenceQ4_K_M · 21.4 GB · ~37 tok/s · 16K24 GB tier →
Mac 32 GB · Mac Mini M4Q4_K_M · 21.4 GB · ~5 tok/s · 32KMacs →
Mac 64 GB · Mac Studio M4 MaxQ4_K_M · 21.4 GB · ~21 tok/s · 32KMacs →

Alternatives to Underdog Saluki 27B 1.0

Similar models

6
ModelScoreVRAMtok/sWhy it is similar
Qwen3.8 27BAlibaba121.517.9 GB~38same size
Qwen3.6 27BAlibaba111.417.9 GB~38same size
Qwen3.5 27BAlibaba103.817.9 GB~38same size
Gemma 3 27B ITGoogle82.421.7 GB~36same size
Gemma 2 27B ITGoogle71.419.5 GB~37same size
Gemma 4 26B A4BGoogle—16.9 GB~153same size

Specifications

Publisher
ConwayResearch
Parameters
27B
Release Date
2026-10-08
License
Apache 2.0

Resources

Frequently Asked Questions

How much VRAM does Underdog Saluki 27B 1.0 need?

Underdog Saluki 27B 1.0 requires 17.8 GB of VRAM at Q4_K_M, or 59.4 GB at BF16.

VRAM = Weights + KV Cache + Overhead

Weights = 27B × 4.8 bits ÷ 8 = 16.2 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.

VRAM usage by quantization

17.8 GB

Learn more about VRAM estimation →

Can NVIDIA GeForce RTX 4090 run Underdog Saluki 27B 1.0?

Yes, at Q5_K_M (21.2 GB) or lower. Higher quantizations like Q6_K (24.5 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.

What's the best quantization for Underdog Saluki 27B 1.0?

For Underdog Saluki 27B 1.0, Q4_K_M (17.8 GB) offers the best balance of quality and VRAM usage. Q5_K_M (21.2 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 12.6 GB.

VRAM requirement by quantization

Q2_K
12.6 GB
Q4_K_M ★
17.8 GB
Q5_K_M
21.2 GB
Q6_K
24.5 GB
Q8_0
29.7 GB
BF16
59.4 GB

★ Recommended — best balance of quality and VRAM usage.

Learn more about quantization →

Can I run Underdog Saluki 27B 1.0 on a Mac?

Yes, but only at lower quantizations. The smallest Mac that can run Underdog Saluki 27B 1.0 is MacBook Pro 14" M3 Pro (18 GB) at Q2_K; 109 of the 156 Macs we list can run it at some quantization. For Q4_K_M (17.8 GB) you need a Mac with more unified memory.

Can I run Underdog Saluki 27B 1.0 locally?

Yes — Underdog Saluki 27B 1.0 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

How fast is Underdog Saluki 27B 1.0?

At Q4_K_M, Underdog Saluki 27B 1.0 can reach ~269 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~37 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 ÷ 17.8 × 0.65 = ~292 tok/s

Estimated speed at Q4_K_M (17.8 GB)

~292 tok/s
~37 tok/s
~292 tok/s
~269 tok/s

Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.

Learn more about tok/s estimation →

What's the download size of Underdog Saluki 27B 1.0?

At Q4_K_M, the download is about 16.20 GB. The full-precision BF16 version is 54.00 GB. The smallest option (Q2_K) is 11.47 GB.

Which GPUs can run Underdog Saluki 27B 1.0?

8 consumer GPUs can run Underdog Saluki 27B 1.0 at Q4_K_M. Top options include NVIDIA GeForce RTX 5090, AMD Radeon RX 7900 XT, AMD Radeon RX 7900 XTX. 1 GPU have plenty of headroom for comfortable inference.

Which devices can run Underdog Saluki 27B 1.0?

125 devices with unified memory can run Underdog Saluki 27B 1.0 at Q4_K_M (17.8 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Asus ROG Flow Z13 (2025, Ryzen AI Max+ 395, 128 GB), Beelink GTR9 Pro (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.