← Home

Stack Builder

Eight inputs — use case, budget, scale, privacy posture — and we compose the full rig: GPU + runtime + 1-3 model picks + first-run workflow + cost rollup + ready-to-paste install script. Three tiers side-by-side so the upgrade path stays visible.

Every recommendation references rule-based scoring; measured tok/s carries a confidence chip when surfaced. We don't invent numbers — when the data isn't there we say so.

Tell us about your build

URL updates as you change fields — share or bookmark a result.

Side-by-side: budget vs balanced vs stretch

one step down · your inputs · one step up
Budget
~$1,500

One step down on budget. What you give up; what you keep.

GPU
NVIDIA RTX PRO 4000 Blackwell· 24 GB
Runtime
Ollama
Top model
Qwen 3 14B· Q4_K_M
3-yr TCO
$1,559
Break-even
317 mo vs cloud
Balanced
~$1,500

Your inputs, our recommendation. Read the full card below.

GPU
NVIDIA RTX PRO 4000 Blackwell· 24 GB
Runtime
Ollama
Top model
Qwen 3 14B· Q4_K_M
3-yr TCO
$1,559
Break-even
317 mo vs cloud
Stretch
~$2,600

One step up on budget. What you'd gain; what it costs.

GPU
NVIDIA RTX PRO 4500 Blackwell· 32 GB
Runtime
Ollama
Top model
Qwen 3 32B· Q4_K_M
3-yr TCO
$2,684
Break-even
423 mo vs cloud

Your recommended stack

full breakdown — read top to bottom
Balanced — recommended
nvidia24 GB VRAM~$1,500

NVIDIA RTX PRO 4000 Blackwell + Ollama + Qwen 3 14B

§ Hardware
32B Q4 — estimated starting point; check artifact size and context
Expected throughput: No matched measurement is used for this recommendation. Check the evidence below for the tested model, quantization and runtime.
Estimated(rule-based scoring)Full hardware page →
§ Runtime
Ollama

Default pick for most operators: one binary, automatic GPU detection, OpenAI-compatible HTTP API at `:11434`. Sufficient for solo + small-team workloads.

  • ›Install: curl -fsSL https://ollama.com/install.sh | sh
  • ›Pull a model: ollama pull <model>:<tag>
  • ›HTTP API at http://localhost:11434
§ Model picks (2)
  • Qwen 3 14B
    14B params
    Q4_K_M
    ~8.6 GB

    Strongest general-purpose model at 14B in 2026. Multilingual tokenizer (1.7× more efficient on Turkish/Asian languages than Llama). Reasoning mode available.

    C
    Community-reported·30-45 tok/s on 16GB VRAM
  • Phi-4 14B
    14B params
    Q4_K_M
    ~8.5 GB

    Microsoft's reasoning-focused 14B trained on heavy synthetic data. Beats Llama 3.1 8B on math/code benchmarks. Weaker creative writing.

    Ed
    Editorial·30-45 tok/s on 16GB VRAM
§ First-run workflow
✓ Curated stack match
Your inputs match the editorial stack Build a local coding-agent stack (May 2026). That page has the field-tested version of this recipe with concrete commands and a why-not-the-alternative for each pick.
  1. Install Ollama on Linux.
  2. Pull the primary model: ollama pull qwen3-14b
  3. Verify it runs: ollama run qwen3-14b — type a test prompt.
  4. Connect a coding agent: install Cline (VS Code extension) or Aider (CLI) and point it at http://localhost:11434 (Ollama) or :8000 (vLLM).
  5. Verify the full loop end-to-end before adding observability, monitoring, or any second model.
§ Total cost of ownership (3-year)
Upfront
$1,500
hardware
Monthly electricity
$2
at 0.34 kWh/day
3-year total
$1,559
upfront + electricity
Cloud equivalent
$229
same token volume
Break-even
317 mo
local beats cloud
Tok/s assumed
80.6
extrapolated
Assumptions: 4 hr/day active, 60% utilization, 3-year amortization, $0.30/M token cloud equivalent. Tune in cost calculator →

Install script

copy and paste — this gets you to first token
#!/usr/bin/env bash
# RunLocalAI stack installer — generated by /stack-builder
# Use case: coding
# Hardware: NVIDIA RTX PRO 4000 Blackwell
# Runtime: Ollama

set -e

# 1. Install Ollama
curl -fsSL https://ollama.com/install.sh | sh

# 2. Pull the primary model
ollama pull qwen3-14b

# 3. Sanity check
ollama run qwen3-14b "Hello — respond in one short sentence to confirm you're running."

# 4. Verify the HTTP API
curl http://localhost:11434/api/tags

Where to go from here

GPU chooser →

Just the hardware-pick question, with side-by-side compare, price/perf scatter, and score breakdown per dimension.

Custom build engine →

Reverse direction: I have this hardware — what fits? Use this to validate the recommendation against your actual rig.

Quant Advisor →

Drill into the model picks: Q4_K_M vs Q5_K_M vs Q8 on your specific VRAM, with quality curve + VRAM fit visualization.

TCO calculator →

Tune every assumption: utilization, electricity rate, cloud equivalent rate, amortization horizon.

Curated stacks →

18 hand-curated stack recipes for specific outcomes (coding agent, offline RAG, dual-3090, Mac cluster, iPhone, etc.)

Apps directory →

37 curated apps that plug into the runtime + model picks above: chat UIs, coding agents, RAG, voice, image, browser ext, mobile, editor plugins. Filter by your VRAM + privacy posture.