LLMBOTTLENECK.COM
Meta / llama / file from the unsloth mirror

Llama-3.2-1B-Instruct

1.24 billion parameters.

Open in the calculator →

Architecturepublished data
Quick answer
Llama-3.2-1B-Instruct needs about 2.02 GB at Q4_K_M — the format most people download — with 8,192 tokens of context. It fits entirely on 18 of 18 common devices, starting with the NVIDIA GeForce RTX 5070 (12 GB). What else fits in 12 GB. No hardware for it? Rent a GPU that runs it, priced live.

Sized at 8,192 tokens of context with the whole model in device memory, across 18 common devices. Speeds are estimates, not benchmarks.

Find the NVIDIA GeForce RTX 5070, 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.

Configuration read from a mirror of a restricted repository

The official repository is meta-llama/Llama-3.2-1B-Instruct (licence: llama3.2). Meta's repository is licence-gated, so its config cannot be read anonymously. This configuration is read from Unsloth's ungated copy, which declares meta-llama/Llama-3.2-1B-Instruct as its base model; it is not byte-compared with Meta's file.

Weight-file size by format

FormatSizeRangeBasis
FP163.00 GB2.47 GB – 3.02 GBreconstructed size
Q8_01.60 GB1.31 GB – 2.09 GBreconstructed size
Q6_K1.23 GB1.01 GB – 1.85 GBreconstructed size
Q5_K_M1.09 GB0.87 GB – 1.74 GBreconstructed size
Q5_01.07 GB0.85 GB – 1.73 GBreconstructed size
Q4_K_M0.95 GB0.73 GB – 1.64 GBreconstructed size
Q4_00.92 GB0.70 GB – 1.60 GBreconstructed size
Q3_K_M0.81 GB0.59 GB – 1.53 GBreconstructed size
Q2_K0.66 GB0.43 GB – 1.40 GBreconstructed size

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

Q4_K_M at 8,192 tokens · 18 of 18 devices hold it entirely
DeviceMemoryVerdictDecodeCalculator
NVIDIA GeForce RTX 50702.02 GB of 12 GBfits~427 tok/sOpen →
AMD Radeon™ RX 9070 XT2.02 GB of 16 GBfits~466 tok/sOpen →
NVIDIA GeForce RTX 40802.02 GB of 16 GBfits~456 tok/sOpen →
NVIDIA GeForce RTX 50802.02 GB of 16 GBfits~610 tok/sOpen →
NVIDIA GeForce RTX 5070 Ti2.02 GB of 16 GBfits~570 tok/sOpen →
NVIDIA GeForce RTX 5060 Ti 16GB2.02 GB of 16 GBfits~285 tok/sOpen →
NVIDIA GeForce RTX 40902.02 GB of 24 GBfits~641 tok/sOpen →
AMD Radeon™ RX 7900 XTX2.02 GB of 24 GBfits~699 tok/sOpen →
NVIDIA GeForce RTX 30902.02 GB of 24 GBfits~595 tok/sOpen →
NVIDIA GeForce RTX 50902.02 GB of 32 GBfits~1139 tok/sOpen →
NVIDIA RTX 6000 Ada Generation2.02 GB of 48 GBfits~610 tok/sOpen →
Apple M4 Pro2.02 GB of 64 GBfits~101 tok/sOpen →
Apple M5 Pro2.02 GB of 64 GBfits~106 tok/sOpen →
AMD Ryzen AI Max+ 395 with Radeon 8060S2.02 GB of 128 GBfits~187 tok/sOpen →
Apple M3 Max2.02 GB of 128 GBfits~117 tok/sOpen →
Apple M4 Max2.02 GB of 128 GBfits~129 tok/sOpen →
NVIDIA DGX Spark2.02 GB of 128 GBfits~174 tok/sOpen →
Apple M2 Ultra2.02 GB of 192 GBfits~141 tok/sOpen →

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 Llama-3.2-1B-Instruct

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
28 configurations in stock · 5 out of stock
MachineSpeedPer hourPer M tokensWhereRent
1× RTX 306012 GBrecommendedcheapestbest value~228 tok/sFaster than you read$0.036$0.043Vast.aiMarketplace · 99.8% reliableRent on Vast.ai
1× RTX 308010 GB~483 tok/sFaster than you read$0.082$0.047Vast.aiMarketplace · 98.3% reliableRent on Vast.ai
1× RTX 30708 GB~284 tok/sFaster than you read$0.082$0.080Vast.aiMarketplace · 99.1% reliableRunPod $0.13/h Rent on Vast.ai
1× RTX 4060 Ti 16GB16 GB~183 tok/sFaster than you read$0.090$0.14Vast.aiMarketplace · 99.4% reliableRent on Vast.ai
1× RTX 407012 GB~320 tok/sFaster than you read$0.096$0.083Vast.aiMarketplace · 99.8% reliableRent on Vast.ai
1× RTX 309024 GB~595 tok/sFaster than you read$0.12$0.057Vast.aiMarketplace · 99.5% reliableRunPod $0.50/h Rent on Vast.ai
1× RTX 5060 Ti 16GB16 GB~284 tok/sFaster than you read$0.12$0.12Vast.aiMarketplace · 99.1% reliableRent on Vast.ai
1× RTX 3080 Ti12 GB~579 tok/sFaster than you read$0.14$0.065Vast.aiMarketplace · 99.9% reliableRent on Vast.ai

No GPU at all

Use Llama-3.2-1B-Instruct by the token

For one person chatting, that is about 2.2× cheaper than the best-value rented card above ($0.043 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
  1. 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.
  2. 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.
  3. 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.
  4. 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 it

No Q4_K_M file of Llama-3.2-1B-Instruct is catalogued here, so there is no exact command to vouch for. Start the machine from the llama.cpp server image ghcr.io/ggml-org/llama.cpp:server-cuda and point -hf at a Q4_K_M build from the publisher or a quantizer you trust — search Hugging Face for one. Check its file size against the memory figure on this page before you rent.

Every machine holds the whole model at Q4_K_M and 8,192 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 Llama-3.2-1B-Instruct 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 read from the mirror's config.json

Retrieved 2026-09-13 at pinned commit 5a8abab4a5d6; parameter count from the safetensors index at the same revision.

Architecture
llama
Layers
16
Hidden size
2,048
Attention heads
32
KV heads
8
Head dimension
64
Feed-forward width
8,192
Vocabulary
128,256
Context ceiling
131,072
RoPE theta
500,000

Where the memory goes

tokenembedding× 16 decoder blocksGrouped-query attention32 query · 8 KV headsfull contextFeed-forwardone networkall activeoutputprojectiongrows with contextfixed per token

Grouped-query attention shares each key/value head across 4 query heads, so the KV cache is 25% of what multi-head attention would need at the same context.

Open the exact config.json this page was read from (the mirror) ↗

More from Meta

Llama-3.2-3B-InstructLlama-3.3-70B-InstructLlama-4-Maverick-17B-128E-InstructLlama-4-Scout-17B-16E-InstructMeta-Llama-3.1-70B-Instruct

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":"unsloth-llama-3-2-1b-instruct","quantization":"Q4_K_M","context":8192,"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="unsloth-llama-3-2-1b-instruct" data-quantization="Q4_K_M"
  data-hardware="nvidia-geforce-rtx-4090" data-context="8192"></script>

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

Citing this page

LLM Bottleneck. “Llama-3.2-1B-Instruct VRAM and hardware requirements.” Architecture from unsloth/Llama-3.2-1B-Instruct at revision 5a8abab4a5d6, retrieved 2026-09-13. https://llmbottleneck.com/models/unsloth-llama-3-2-1b-instruct

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

llmbottleneck
catalogue 2026-10-03models 327devices 135