Undi95 / llama

ReMM-SLERP-L2-13B

Name: ReMM-SLERP-L2-13B architecture and memory requirements
Creator: LLM Bottleneck
License: https://creativecommons.org/licenses/by/4.0/

Parameter count not published.

Open in the calculator →

Architecturepublished data

Quick answer

ReMM-SLERP-L2-13B needs about 18.0 GB at Q8_0 with 4,096 tokens of context. It fits entirely on 12 of 18 common devices, starting with the NVIDIA GeForce RTX 4090 (24 GB). What else fits in 24 GB. No hardware for it? Rent a GPU that runs it, priced live.

Smallest device that fitsNVIDIA GeForce RTX 409024 GB · 1008 GB/s~40 tok/s Faster than you readOpen in the calculator →
Fastest in this listNVIDIA GeForce RTX 509032 GB · 1792 GB/s~72 tok/s Faster than you readOpen in the calculator →

Sized at 4,096 tokens of context with the whole model in device memory, across 18 common devices. Speeds are estimates, not benchmarks.

Find the NVIDIA GeForce RTX 4090, the smallest card listed here that holds it: Amazon ↗ · eBay (new and used) ↗Store links may pay us a commission. They never decide which card is suggested — the memory arithmetic does.

Weight-file size by format

every format, with its evidence

Format	Size	Range	Basis
Q8_0	13.8 GB	exact	published data
Q6_K	10.7 GB	exact	published data
Q5_K_M	9.23 GB	exact	published data

3 of these are published files. Open the pinned GGUF repository ↗

A file size is not the memory a run needs: the KV cache and the runtime reserve come on top, and the calculator adds both.

What the labels mean

published data: Read from a published source — a file's byte count, a model's configuration or a manufacturer's specification — or exact arithmetic on such values. Not a measurement on a machine.
reconstructed size: Weight size reconstructed from the pinned architecture, because no published file exists.
size range: Only a lower and an upper bound are claimed for this weight size.

Hardware ladder

Q8_0 at 4,096 tokens · 12 of 18 devices hold it entirely

Device	Memory	Verdict	Decode	Calculator
NVIDIA GeForce RTX 5070	18.0 GB of 12 GB	18 layers on system RAM	~7.9 tok/s with offload	Open →
AMD Radeon™ RX 9070 XT	18.0 GB of 16 GB	6 layers on system RAM	~16 tok/s with offload	Open →
NVIDIA GeForce RTX 4080	18.0 GB of 16 GB	6 layers on system RAM	~15 tok/s with offload	Open →
NVIDIA GeForce RTX 5080	18.0 GB of 16 GB	6 layers on system RAM	~18 tok/s with offload	Open →
NVIDIA GeForce RTX 5070 Ti	18.0 GB of 16 GB	6 layers on system RAM	~17 tok/s with offload	Open →
NVIDIA GeForce RTX 5060 Ti 16GB	18.0 GB of 16 GB	6 layers on system RAM	~12 tok/s with offload	Open →
NVIDIA GeForce RTX 4090	18.0 GB of 24 GB	fits	~40 tok/s	Open →
AMD Radeon™ RX 7900 XTX	18.0 GB of 24 GB	fits	~44 tok/s	Open →
NVIDIA GeForce RTX 3090	18.0 GB of 24 GB	fits	~37 tok/s	Open →
NVIDIA GeForce RTX 5090	18.0 GB of 32 GB	fits	~72 tok/s	Open →
NVIDIA RTX 6000 Ada Generation	18.0 GB of 48 GB	fits	~38 tok/s	Open →
Apple M4 Pro	18.0 GB of 64 GB	fits	~14 tok/sAbout reading pace	Open →
Apple M5 Pro	18.0 GB of 64 GB	fits	~15 tok/sAbout reading pace	Open →
AMD Ryzen AI Max+ 395 with Radeon 8060S	18.0 GB of 128 GB	fits	~12 tok/sAbout reading pace	Open →
Apple M3 Max	18.0 GB of 128 GB	fits	~20 tok/sAbout reading pace	Open →
Apple M4 Max	18.0 GB of 128 GB	fits	~26 tok/sAbout reading pace	Open →
NVIDIA DGX Spark	18.0 GB of 128 GB	fits	~11 tok/sAbout reading pace	Open →
Apple M2 Ultra	18.0 GB of 192 GB	fits	~36 tok/s	Open →

Decode is a calibrated estimate from the published calibration; capacity is the manufacturer's published ceiling, not guaranteed free memory. Fitting in memory is not the same as loading: whether the runtime and version you have supports this architecture and format on that machine has not been tested here. Offloaded rows assume there is enough system RAM for the overflow — the calculator checks that against the RAM you declare. Each “Open” link keeps this model, format and context.

Run it in the cloud

Rent a GPU that runs ReMM-SLERP-L2-13B

Every rentable machine that holds the whole model, from Vast.ai and RunPod, with this site's speed estimate for each and the provider's own price, read live. Cheapest first; sort by cost per token or speed instead, or switch the format.

Prices · 18:30 UTC, 22 Sept

Machine	Speed	Per hour	Per M tokens	Where	Rent
2× RTX 306024 GBcheapest	~28 tok/sAbout reading pace	$0.11	$1.05	Vast.aiMarketplace · 98.0% reliable	Rent on Vast.ai
1× RTX 309024 GBrecommendedbest value	~37 tok/sFaster than you read	$0.12	$0.91	Vast.aiMarketplace · 99.5% reliableRunPod $0.50/h	Rent on Vast.ai
1× RTX 3090 Ti24 GB	~40 tok/sFaster than you read	$0.19	$1.29	Vast.aiMarketplace · 99.6% reliableRunPod $0.27/h	Rent on Vast.ai
2× RTX 407024 GB	~40 tok/sFaster than you read	$0.19	$1.31	Vast.aiMarketplace · 99.8% reliable	Rent on Vast.ai
2× RTX 4060 Ti 16GB32 GB	~23 tok/sAbout reading pace	$0.22	$2.71	Vast.aiMarketplace · 99.7% reliable	Rent on Vast.ai
1× L424 GB	~12 tok/sAbout reading pace	$0.26	$5.92	Vast.aiMarketplace · 98.9% reliableRunPod $0.49/h	Rent on Vast.ai
2× RTX 5060 Ti 16GB32 GB	~35 tok/sFaster than you read	$0.26	$1.98	Vast.aiMarketplace · 99.5% reliable	Rent on Vast.ai
1× RTX 6000 Ada48 GB	~38 tok/sFaster than you read	$0.26	$1.89	Vast.aiMarketplace · 99.9% reliableRunPod $0.84/h	Rent on Vast.ai

No GPU at all

Use ReMM-SLERP-L2-13B by the token

OpenRouter$0.65 / M output · $0.35 / M inputOpen on OpenRouter

For one person chatting, that is about 1.4× cheaper than the best-value rented card above ($0.91 per million tokens). A rented GPU pays off when you keep it busy — many requests batched together, long agent runs — or when the data must stay on a machine you control, or the exact file you want is not served anywhere.

First time renting a GPU? How it works, in four steps

Create an account and add credit. Both providers are prepaid: $10 is enough to try any card on this page for hours. A new RunPod account that arrives through a referral link gets a one-time $5 credit when it first adds $10.
Pick the card and the count shown here, and a template: llama.cpp or vLLM for an OpenAI-compatible API, or one with a web chat if you only want to talk to the model. “How to launch it” below gives the exact image and command.
Wait for the download. The weights are fetched on the machine; tens of gigabytes take a few minutes on a data-centre connection. You pay by the second from the moment it starts.
Stop it when you are done. A running machine bills even when idle. A stopped one costs nothing per hour, though its disk is usually still billed until you delete it.

How to launch Q8_0 on it (llama.cpp)

Container image: ghcr.io/ggml-org/llama.cpp:server-cuda
Start command / arguments: -hf Undi95/ReMM-SLERP-L2-13B-GGUF:Q8_0 -c 4096 -ngl 999 --host 0.0.0.0 --port 8080
Expose port: 8080 — an OpenAI-compatible API at /v1

Starting from a saved llama.cpp template instead? Leave the arguments empty and set these variables — only the first two change between models:

LLAMA_ARG_HF_REPO=Undi95/ReMM-SLERP-L2-13B-GGUF:Q8_0
LLAMA_ARG_CTX_SIZE=4096
LLAMA_ARG_N_GPU_LAYERS=999
LLAMA_ARG_HOST=0.0.0.0
LLAMA_ARG_PORT=8080

Already have the runtime? llama-server -hf Undi95/ReMM-SLERP-L2-13B-GGUF:Q8_0 -c 4096 -ngl 999 --host 0.0.0.0 --port 8080. Weights: Undi95/ReMM-SLERP-L2-13B-GGUF. The CUDA image is for NVIDIA cards; an AMD machine needs llama.cpp's ROCm build.

Every machine holds the whole model at Q8_0 and 4,096 tokens of context, no offload. Speeds are this site’s single-stream decode estimates; prices are what each provider’s own API quoted, on-demand, for the whole machine. Vast hosts below 98% measured reliability are left out.

Referral links Vast.ai, RunPod and Novita pay us a share of what you spend if you sign up through these buttons. It costs you nothing, and it never decides an order or a recommendation: both are computed from the live price and the speed, and options that pay us nothing are listed and recommended on the same terms. How we rank

Buying a card instead, or paying by the token? Compare a year of ReMM-SLERP-L2-13B three ways →

Under the hood

Architecture, read from the publisher’s file

The numbers every figure above is computed from, with the file they came from.

Architecture

architecture.json · pinned 9cf491450795

✓ Architecture read from the published config.json

Retrieved 2026-09-01 at pinned commit 9cf491450795.

Architecture: llama
Layers: 40
Hidden size: 5,120
Attention heads: 40
KV heads: 40
Feed-forward width: 13,824
Vocabulary: 32,000
Context ceiling: 4,096

Where the memory goes

Every attention head caches its own keys and values, so the KV cache grows at the full multi-head rate with context.

Open the exact official config.json ↗

Use this answer in your own product

As JSON, from the API

curl -s https://llmbottleneck.com/v1/analyze \
  -H "Authorization: Bearer $LLMB_KEY" -H "content-type: application/json" \
  -d '{"model":"undi95-remm-slerp-l2-13b","quantization":"Q8_0","context":4096,"hardware":"nvidia-geforce-rtx-4090"}'

Same engine, same evidence, every field sourced. Free key in one step →

On your page, as a widget

<script src="https://llmbottleneck.com/widget.js"
  data-model="undi95-remm-slerp-l2-13b" data-quantization="Q8_0"
  data-hardware="nvidia-geforce-rtx-4090" data-context="4096"></script>

No key needed. Unbranded, with your own buy button, on Pro and Business →

Citing this page

LLM Bottleneck. “ReMM-SLERP-L2-13B VRAM and hardware requirements.” Architecture from Undi95/ReMM-SLERP-L2-13B at revision 9cf491450795, retrieved 2026-09-01. https://llmbottleneck.com/models/undi95-remm-slerp-l2-13b

Every figure above is either the published value or a reconstruction whose measured error is on the accuracy page.