BigCode15B~10 GB VRAM at Q4_K_M

StarCoder 2 15B — VRAM, speed & local setup

Written by Jakub Rusinowski · Last updated

The flagship StarCoder model. HumanEval score of 46.4%, beating many larger general-purpose models on code tasks. Excellent at generating and explaining complex code in 600+ languages.

StarCoder 2 15B needs about 10 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache. On Apple Silicon that figure comes out of unified memory.

Logic78
Creative48
Coding94

VRAM and speed by quantization

Quoted against NVIDIA RTX 4090 (24 GB). Includes the KV cache at 8K context, so it reads higher than the headline figure.

QuantVRAMSpeed (est.)Fit
Q2_K
2.63 bpw
6.6 GB~106 tok/sFits
Q3_K_M
3.41 bpw
8.1 GB~87 tok/sFits
Q4_K_M
4.83 bpw
10.8 GB~66 tok/sFits
Q5_K_M
5.67 bpw
12.5 GB~58 tok/sFits
Q6_K
6.56 bpw
14.2 GB~51 tok/sFits
Q8_0
8.50 bpw
17.9 GB~41 tok/sFits
F16
16.00 bpw
32.5 GB~3 tok/sOffload

Black marker = usable memory on the NVIDIA RTX 4090 (24 GB). Estimates from the memory-bandwidth roofline on the methodology page. StarCoder 2 15B VRAM calculator →

Get StarCoder 2 15B running

The cheapest catalogued GPU that runs StarCoder 2 15B is the Intel Arc B580 (12 GB).

Buy this hardware Intel Arc B580 12GB — 12 GB VRAM · 190 W board powerDeploy in the cloud now RTX 4090 on RunPod — from $0.34/hr · rate checked 2026-07

or compare on Vast.ai from $0.35/hr (typical low · varies)

As an Amazon Associate we earn from qualifying purchases. Cloud GPU links are referral links — we may earn a commission at no extra cost to you.

Affiliate disclosure: Some links on this page are affiliate links — if you buy through them, LLM Configurator may earn a commission at no extra cost to you. As an Amazon Associate, LLM Configurator earns from qualifying purchases.
Intel Arc B580 12GB
12 GB VRAM · 190 W board power
2026 prices are volatile — check the current listing.

How to run StarCoder 2 15B

Install Ollama, then run:

ollama run starcoder2:15b
Weights on Hugging Face: bigcode/starcoder2-15b ↗

Specifications

Curated — A hand-written entry from before this catalogue recorded its sources. The figures are long-standing but their provenance is not on file.

Parameters
15 Billion
Context window
16,384
Architecture
Dense
Provider
BigCode
Licence
BigCode Open RAIL-M v1
Practical deployment
Laptop
Specified at
Q4_K_M
System RAM
24 GB
Record updated
2024-02-28
LicenceBigCode OpenRAIL-MCommercial use permitted

Weights are downloadable and commercial use is permitted, subject to the licence’s acceptable-use terms.

Quality and use cases

Scores as published by the model’s authors or an independent evaluator — quality, not throughput, and not measured by us.

Best forcoding agentcode reviewdev work

Can I run StarCoder 2 15B on my GPU?

Other StarCoder 2 sizes

StarCoder 2 15B — frequently asked questions

How much VRAM does StarCoder 2 15B need?

About 10 GB at Q4_K_M — quantized weights plus framework overhead, before any KV cache. The cache grows with context length and is added on top; the table above folds it in. Apple Silicon counts unified memory toward the same figure.

Does StarCoder 2 15B run on an RTX 4090 (24 GB)?

Yes. StarCoder 2 15B needs about 10 GB at Q4_K_M, inside a 24 GB card, at an estimated 66 tokens/sec.

How do I run StarCoder 2 15B locally?

Install Ollama and run `ollama run starcoder2:15b`. That pulls the weights and starts a local OpenAI-compatible endpoint; after the download nothing leaves the machine.

What other sizes does StarCoder 2 come in?

StarCoder 2 3B (3 GB), StarCoder 2 7B (5 GB), StarCoder 2 15B (10 GB). Every size shares the family's training and licence; the larger ones score higher and need proportionally more memory.