RUNLOCALAIv38
→WILL IT RUNBEST GPUCOMPARETROUBLESHOOTSTARTPULSEMODELSHARDWARETOOLSBENCH
RUNLOCALAI

Independently operated catalog for local-AI hardware and software. Hand-written verdicts. Source-cited claims. Reproducible commands when we have them.

OP·Fredoline Eruo
DIR
  • Models
  • Hardware
  • Tools
  • Benchmarks
TOOLS
  • Will it run?
  • Compare hardware
  • Cost vs cloud
  • Choose my GPU
  • Quick answers
REF
  • All buyer guides
  • Methodology
  • Glossary
  • Errors KB
  • Trust
EDITOR
  • About
  • Author
  • How we make money
  • Editorial policy
  • Contact
LEGAL
  • Privacy
  • Terms
  • Sitemap
MAIL · MONTHLY DIGEST
Get monthly local AI changes
Monthly recap. No spam.
DISCLOSURE

Some links on this site are affiliate links (Amazon Associates and other first-class retailers). When you buy through them, we earn a small commission at no extra cost to you. Affiliate links do not influence our verdicts — there are cards we rate highly that we don't have affiliate relationships with, and cards that sell well that we refuse to recommend. Read more →

© 2026 runlocalai.coIndependently operated
RUNLOCALAI · v38
← Back to Will-it-run

Custom build engine

Describe your build — any GPUs, CPU, RAM, OS, runtime, use case. We'll compute effective VRAM honestly, recommend a runtime, and tell you which models fit comfortably, which are borderline, and which aren't practical.

Total VRAM ≠ pooled VRAM. We never sum VRAM unless the silicon truly pools (Apple unified memory). We always explain why effective is lower than total.

Describe your build

Add GPUs, set CPU/RAM/OS, optionally pick a runtime + use case. URL updates as you change fields — share a build by copying the URL.

Build summary

Total VRAM
13824 GB
Effective VRAM
~13823 GB
range 12716-13822 GB
Topology
single gpu
none
Setup difficulty
beginner
speed penalty ~0%
Why effective VRAM is lower than total

Single NVIDIA GB200 NVL72 — 13824 GB VRAM minus ~1.5 GB runtime overhead = ~13822 GB usable for weights + KV cache + activations. The 8% headroom we reserve covers the typical OS/driver footprint and gives KV-cache room for an 8K-32K context.

Coding agents — agentic tool-call burst

Workload-specific bottleneck. Where this kind of work actually breaks first, and what to budget for.

Bottleneck: kv cache

Coding agents emit 5-15 tool calls per task. Each call carries the full agent system prompt + context. KV-cache budget for that prompt × concurrent requests is the limit. The decode side is well-served by any modern card; the prefill side bottlenecks first.

Budget for
  • •32K context with KV-cache room to spare (~3-4 GB on 4090 AWQ-INT4)
  • •Prefix cache: prefer SGLang for >5 tool calls / task
  • •Decode latency: aim for >40 tok/s sustained

Recommended runtime

Best engine for this topology + skill level + use case.

vLLM
primary
involved

AWQ-INT4 path fits 32B-class models on a 24 GB card with concurrent users. The production-default for self-hosted coding agents and multi-user serving.

ExLlamaV2
alternative
involved

Single-stream throughput king on consumer NVIDIA. EXL2 4.65bpw on a 4090 hits the highest tok/s in this class.

Models that fit your build

47 models considered (filtered by coding). Categorized by headroom at the recommended quant + a sensible context for your use case.

Comfortable
24 models · ≥15% headroom
ModelParamsQuantVRAM est.ContextNote
DeepSeek V4 Pro (1.6T MoE)1600BQ4_K_M4248.9 GB32,768Comfortable fit with 69% headroom — room to extend context or run alongside other workloads.
Kimi K2.61000BQ4_K_M2668.5 GB32,768Comfortable fit with 81% headroom — room to extend context or run alongside other workloads.
Ring-2.6-1T1000BFP164134.6 GB32,768Comfortable fit with 70% headroom — room to extend context or run alongside other workloads.
DeepSeek V4745BAWQ-INT42306.8 GB32,768Comfortable fit with 83% headroom — room to extend context or run alongside other workloads.
DeepSeek V4 Flash (284B MoE)284BQ5_K_M807.6 GB32,768Comfortable fit with 94% headroom — room to extend context or run alongside other workloads.
DeepSeek Coder V2 236B236BQ4_K_M656.2 GB32,768Comfortable fit with 95% headroom — room to extend context or run alongside other workloads.
DeepSeek V2.5 236B236BQ4_K_M656.2 GB32,768Comfortable fit with 95% headroom — room to extend context or run alongside other workloads.
Qwen 3 235B-A22B235BQ5_K_M674.2 GB32,768Comfortable fit with 95% headroom — room to extend context or run alongside other workloads.
Llama 4 Scout109BFP16481.5 GB32,768Comfortable fit with 97% headroom — room to extend context or run alongside other workloads.
Qwen 3 72B72BAWQ-INT4254.2 GB32,768Comfortable fit with 98% headroom — room to extend context or run alongside other workloads.
Llama 4 70B70BAWQ-INT4248.1 GB32,768Comfortable fit with 98% headroom — room to extend context or run alongside other workloads.
Llama 3.3 70B Instruct70BQ8_0123.4 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
Llama 3.1 70B Instruct70BQ5_K_M225.1 GB32,768Comfortable fit with 98% headroom — room to extend context or run alongside other workloads.
Qwen 3.6 35B-A3B (MTP)35BFP16178.1 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
Phind CodeLlama 34B v234BQ4_K_M73.7 GB16,384Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
DeepSeek Coder V333BAWQ-INT4135.2 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
Qwen 2.5 32B Instruct32BQ8_0134.3 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
Qwen 3 32B32BQ8_0134.3 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
Qwen 2.5 Coder 32B Instruct32BQ8_078.9 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
DeepSeek R1 Distill Qwen 3 32B32BAWQ-INT4132.2 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
Qwen 3 Coder 32B32BAWQ-INT4132.2 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
Gemma 4 31B Dense31BQ8_0131.2 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
Qwen 3 30B-A3B30BQ8_0128 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
Qwen 3.6 27B (MTP)27BFP16145.3 GB32,768Comfortable fit with 99% headroom — room to extend context or run alongside other workloads.
Borderline
0 models · tight, may need quant downgrade

No borderline models — clean fit ladder.

Not practical
0 models · oversize for this build

Every considered model fits.

Related

Multi-GPU buying guide →

NVLink vs PCIe, tensor- vs pipeline-parallel, mixed-card honesty.

Hardware combinations →

Curated multi-GPU / cluster setups with effective-VRAM math.

Setup path-finder →

OS + runtime install commands for your stack.

Compatibility matrix →

Runtime × OS × hardware support truth table.

Shopping a full build instead of a single card?

If you're sizing a fresh AI build (not just a card to drop into an existing system), the build-budget walkthroughs cover the whole BOM honestly: AI PC build under $1,000 or AI PC build under $2,000 cover the realistic 2026 budget tiers.

Vertical-fit shopping? AI PC for students covers the budget + portability tradeoffs; AI PC for developers covers the coding workflow specifics; AI PC for small business covers the document-RAG / always-on machine.

Form-factor first? See best laptop for local AI, best Mac for local AI, best mini PC for local AI, or best used GPU for local AI.

See something off?Submit a benchmark·Report outdated·Suggest a correctionWe read every submission. Editorial review takes 1-7 days.