RUNLOCALAIv38
->Will it run?Best GPUCompareTroubleshootStartLearnPulseModelsHardwareToolsBench
Run check
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
  • Prompting kits
  • Quick answers
REF
  • All buyer guides
  • Learn local AI
  • 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
CATALOG·runlocalai.co|HARDWARE·134|MODELS·315|PULSE·6
v38
CAN I RUN LOCAL AI? · WILL-IT-RUN FRAMEWORK · LIVE|v38 / catalogv38 / catalog

> Will it run on your hardware?

Enter your GPU, RAM, and model target. The RunLocalAI Will-It-Run Framework shows what fits, how fast it should feel, when the number is measured, and what to upgrade next. Honest verdicts, auditable methodology, confidence labels on every benchmark.

134
Hardware
315
Models
39
Bench rows
11+
Tools
->Check my hardwarePick a GPUCompare hardwareView benchmarks
Sample answer
RTX 3060 12GB + 32GB RAM

Llama 3.1 8B Q4 fits comfortably; 70B needs offload or more VRAM. Confidence changes to measured when a benchmark row exists.

Run mine ->
> /state-of-local-ai-2026> /openclaw> /clusters> /macos-26.5> /cost-calculator
NEW HERE?Start with the 12-min primer →·Or copy a Docker bundle →
HARDWARE INDEX
134profiles
MODEL CATALOG
315models
TASK MATRIX
39bench rows
BLK 00 — DECISION

Tell us what you're asking.

All paths →
§ Editorial stance
We're for running AI locally where it makes sense. Cloud APIs from Anthropic, OpenAI, Google, DeepSeek, and the open-weight hosters are world-class — modern local AI is downstream of their research. They're the right call for frontier-quality work, day-one access to new models, and workloads where you don't want to own hardware. Local is the right call for high-volume routine work, privacy-sensitive content, and offline capability. Our calculators are tools, not advocacy — they show you where each option lands for your specific workload so you can pick deliberately, not by default.
39
Measured rows
0
Reproduced
16
Queued gaps
1
Critical gaps
Q.01
Will it run?

Pick your hardware + a model, get a verdict in seconds.

Check fit →
Q.02
How well will it run?

Every hardware unit ranked by RunLocalAI Score (0–1000).

See leaderboard →
Q.03
What should I buy?

Tell us budget + OS + workload, get a GPU recommendation.

Get recommendation →
Q.04
Build me the whole stack

Use case + budget + scale → GPU + runtime + models + install script. Three tiers side-by-side.

Open stack builder →
Q.05
What would I save?

Paste a prompt or code. See exactly what 11 cloud providers would charge vs your local rig. Break-even in months.

Open cost calculator →
Q.06
Which quant should I use?

Q4_K_M vs Q5_K_M vs Q8 settled with math, not folklore. Quality curve + VRAM fit visualization.

Open quant advisor →
Q.07
How fast does it feel?

Watch tokens stream at your hardware's actual speed, side-by-side with Claude / GPT-5 / Groq. Race mode.

Open stream visualizer →
Q.08
What did my session cost?

Paste a Claude Code .jsonl. See per-tool breakdown, cache savings, insights. Privacy-first.

Open decoder →
Q.09
Which model fits this use case?

Pick 2 models. Get a 10-row diff: params, context, license, TPS, fit, popularity. Use-case-weighted better-fit verdict.

Open battle card →
Q.10
When does local pay back?

Drag a tokens/day slider, watch the cloud vs local cost lines extend over your horizon. Crossover point drawn. One screenshot.

Open compounder →
Q.11
What apps work with my rig?

37 curated apps that plug into your local runtime: chat UIs, coding agents, RAG, voice, image, browser, mobile, editor plugins. Filter by VRAM + privacy.

Open apps directory →
BLK 01 — HARDWARE

Editor verdicts — recently updated.

ALL 129 →
№
UNIT
VRAM
MSRP
→
01
AMD Radeon RX 9060 XT
16 GB
$449
→
02
NVIDIA RTX 5000 PRO Blackwell 48GB
48 GB
$5,499
→
03
NVIDIA GeForce RTX 3080 Ti
12 GB
$1,199
→
04
NVIDIA GeForce RTX 3070 Ti
8 GB
$599
→
05
NVIDIA GeForce RTX 3060 Ti
8 GB
$399
→
06
NVIDIA GeForce RTX 3050
8 GB
$249
→
07
NVIDIA GeForce RTX 3050 Ti (Mobile)
4 GB
—
→
08
NVIDIA GeForce RTX 2080 Ti
11 GB
$1,199
→
BLK 02 — MODULES

Subroutines operators run.

GPU.SELECT
Best GPU for local AI in 2026
From $200 used to $40K datacenter — what to actually buy at every tier.
→
COMPAT.PROBE
Will my computer run local AI?
Tell us your hardware. We tell you what runs and how fast.
→
DIFF.4090-5090
RTX 4090 vs RTX 5090
Which one's worth it for local AI in 2026?
→
DIFF.M4-4090
MacBook Pro M4 Max vs RTX 4090
Mac unified memory vs NVIDIA CUDA — honest tradeoffs.
→
TASK.CODE
Best local AI for coding
Self-hosted Copilot. Models, hardware, runtimes.
→
TASK.RAG
Best local AI for private docs
RAG over your sensitive files. Nothing leaves your machine.
→
BLK 03 — EVENT LOG

Live stream of what changed.

FULL FEED →
[001]T-4hRUNTIME
llama.cpp ships b9524
[002]T-4hRUNTIME
llama.cpp ships b9528
[003]T-4hRUNTIME
llama.cpp ships b9529
[004]T-4hRUNTIME
vllm ships v0.22.1
[005]T-15hRUNTIME
ollama ships v0.30.5
[006]T-23hRUNTIME
llama.cpp ships b9515
BLK 04 — MODELS

Open-weight registry.

ALL 315 ->
01all-MiniLM-L6-v2
→
02FLUX.1 [dev]
→
03DeepSeek V4 Pro (1.6T MoE)
→
04Qwen 3.5 235B-A17B (MoE)
→
05Qwen 3 235B-A22B
→
06Qwen 3 0.6B
→
07BGE Large EN v1.5
→
08Llama 3.1 8B Instruct
→
09Nomic Embed Text v1.5
→
10Llama 4 Scout
→
11DeepSeek V4 Flash (284B MoE)
→
12Kokoro 82M
→
BLK 05 — OPERATING PARAMETERS

How this instrument operates.

PARAM.HONEST
Honest

Every verdict has a named author and a real opinion. We say what to skip as much as what to buy. Some links are affiliate; the opinion comes first.

PARAM.REPRODUCIBLE
Reproducible

Benchmarks run with documented commands and open-weight models. Confidence tiers — measured / reproduced / inferred — visible on every result. /trust →

PARAM.LOCAL-FIRST
Local-first

Privacy + cost + control beat cloud APIs for most everyday work. We'll tell you when local is the wrong answer too — /explained →

Get monthly local AI changes

Monthly recap of local-AI changes. No spam, unsubscribe with one click.

CATALOG / PUBLICHARDWARE / 134BENCHMARKS / 39
RUNLOCALAI / v38© 2026