AI & Local LLMs

Local models, AI rigs and the tools that run them

New here? Start with these

Decide with live data

Macro photo of rows of identical blank black memory chips on a circuit board, their pins lit in indigo-violet against a near-black background

Qwen3.8 27B Hardware Requirements: VRAM by Quant, and Why Context Costs So Little

A 24 GB card runs Qwen3.8 27B at Q4 or Q5. Q8 needs 32 GB. The surprise is the context cache: 128K tokens costs about 8 GB, a quarter of what an older 32B model charges.

Voltage · 1d ago · 9 min

Macro photo of a row of plain black memory chips on a circuit board, a streak of indigo-violet and coral light on the traces, near-black background

Got an RTX 5090? The Best Local LLMs for 32GB of VRAM in 2026

32GB does not unlock 70B models. It buys better quants and a much longer memory for the 27B to 35B class. Five picks that fit a 5090, with the math shown.

Chiptune · 3d ago · 10 min

Four small black computer towers with warm-lit vents in a row on a dark desk, a taller server unit with a red status light blurred behind them

Ollama vs llama.cpp vs LM Studio vs vLLM: Pick by Who's Using It

One person on one machine, or a team hitting one server? That single question decides it. Then licence, GPU and Mac support, and the API your app talks to, all read from each project's own docs.

Tensor · 4d ago · 9 min

$Hand-drawn editorial illustration of a robot hand holding a magnifying glass over a long paper scroll of boxes, a few ticked in coral and most still empty$

AI keeps 'solving' famous math problems. Here's what actually checks out (2026)

Olympiad golds, Erdős problems, a disproved 87-year-old conjecture and a better Riemann bound. We sorted every big AI math claim since 2024 by who checked it, and which ones a computer can verify line by line.

Tensor · 4d ago · 11 min

Low-angle photo of a row of blank matte memory modules on a motherboard in the foreground, with a plain graphics card blurred behind them under indigo and coral light

Best MoE Models for a 12GB or 16GB GPU: The --n-cpu-moe Number and the RAM Each Needs

Seven mixture-of-experts models, from gpt-oss 20B to Qwen3 235B, split into what stays on your card and what moves to system RAM. The card barely changes. The RAM you need goes from 16 GB to 96 GB.

Voltage · 4d ago · 10 min

Photo of a plain, unbranded dual-fan graphics card standing upright on a dark desk, a monitor behind it glowing with an abstract painted landscape in indigo-violet and coral

No Nvidia Card? Running Stable Diffusion and FLUX on AMD Radeon and Intel Arc (2026)

Your Radeon or Arc card can make AI images. The trick is picking the right backend, because three of the seven options are dead ends. We read ComfyUI's code, AMD's ROCm docs and Intel's PyTorch guide to sort them, then matched FLUX, SD 3.5 and SDXL to each card size.

Pixel · 4d ago · 10 min

Hand-drawn editorial illustration of loose paper sheets being stitched by a coral thread into a tidy web of linked pages on a near-black background

Karpathy's LLM Wiki: Build a Markdown Second Brain That Your AI Keeps Up to Date

Andrej Karpathy shared an idea file for a notes folder the AI writes and maintains, so you stop re-explaining your documents every time you ask a question. Here is what it says, how it works, and where it can go wrong.

Neon · 4d ago · 8 min

Flat vector illustration of a person in a hoodie sitting cross-legged beside an open filing cabinet drawer of blank documents joined by thin coral lines, on an indigo-violet background

Google's Open Knowledge Format (OKF), Explained: Markdown Files That Agents Can Trust

Google Cloud's new spec for agent knowledge is a folder of markdown files with a few lines of YAML on top. Here is what is in the file, what v0.2 added, and where it differs from a vector index.

Glitch · 4d ago · 8 min

Editorial infographic of three model badges converging on a single 24GB graphics card with a fork in the path, near-black background, indigo-violet with a coral accent

Which 24GB-class model should you run? Mistral Small 24B vs Qwen3.6 27B vs Gemma 4 31B

You've got a 24GB card and three excellent dense models that fit it. Here's how Mistral Small 24B, Qwen3.6 27B and Gemma 4 31B differ on VRAM, context and modality — and a clear pick for each kind of user.

Tensor · Sep 26 · 7 min

Isometric diorama of a code editor window and a graphics card together on a desk with a small padlock, near-black background, indigo-violet with a coral accent

Qwen2.5-Coder-32B locally: a private Copilot on a 24GB card

Qwen2.5-Coder-32B still runs a serious private coding assistant on a single 24GB card, with a 32K native context (128K with YaRN). The VRAM math, the tight-but-real fit, and how it stands against the newer Qwen3-Coder.

Tensor · Sep 26 · 7 min

Realistic painterly illustration of a small compact glowing processor on a dark surface, deep indigo-violet with a warm coral highlight

Phi-4 locally: a small model that's still handy, with one dated catch

Microsoft's Phi-4 is a compact 14B that runs on a mid-range card — but at 16K context and a late-2024 release, it shows its age next to newer 14B-class models. The VRAM math and an honest verdict.

Tensor · Sep 26 · 7 min

Editorial infographic of a solid dense 27B model block fitting neatly inside a 24GB graphics card, near-black background, indigo-violet with a coral accent

Qwen3.6 27B locally: the all-dense 24GB Qwen

Qwen3.6's dense 27B is the no-fuss 24GB-card option — every parameter fires, nothing to offload, a real 256K context. The VRAM math, the GPU, and how it differs from the 35B-A3B MoE.

Tensor · Sep 26 · 7 min

Isometric diorama of a huge grid of expert blocks overflowing off a single graphics card into several system-RAM sticks, near-black background, indigo-violet with a coral accent

Qwen3-Next 80B-A3B locally: an 80B at home, if you can hold it

Qwen3-Next 80B-A3B activates only 3B parameters per token — so it's fast. But 80B of weights still have to live somewhere. The honest VRAM math, the offload trick, and what it really takes.

Tensor · Sep 26 · 7 min

Editorial infographic of a mixture-of-experts model: a few active expert blocks on a GPU and the rest resting in system memory, near-black background, indigo-violet with a coral accent

Gemma 4 26B-A4B locally: the Gemma that runs fast on 16GB

Gemma 4's mixture-of-experts model holds about 26B parameters but activates only 3.8B — so it's quick, and it fits a 16GB card at Q4. The VRAM math, the GPU, and the run command.

Tensor · Sep 26 · 7 min

Realistic painterly illustration of a large glowing processor die radiating soft light on a dark surface, deep indigo-violet with a warm coral highlight

Gemma 4 31B locally: the 24GB-card Gemma

Gemma 4 31B is the big dense Gemma — the one that asks for a 24GB card and rewards it. The VRAM per quant, the GPU that fits, and how to run it without pulling the wrong tag.

Tensor · Sep 26 · 7 min

Isometric diorama of a grid of many small expert blocks with only a few lit up feeding a token, near-black background, indigo-violet with a coral accent

Qwen3.6 35B-A3B locally: a 35B that thinks like a 3B

Qwen3.6's default is a mixture-of-experts model: 35B on disk, 3B active per token. That makes it fast — but the VRAM math has a twist worth understanding before you buy a card.

Tensor · Sep 26 · 7 min

Editorial infographic of a compact AI model badge flowing into a mid-range GPU that comfortably fits it, near-black background, indigo-violet with a coral accent

Qwen3.5 9B locally: the model 'ollama run qwen3.5' actually gives you

Type ollama run qwen3.5 and this is the model you get — the 9B. What it is, how little VRAM it needs, and the cheapest card that runs it well.

Tensor · Sep 26 · 7 min

Isometric diorama of a friendly glowing AI chip seated on a mid-range graphics card, near-black background, indigo-violet with a coral accent

Gemma 4 12B locally: the VRAM, the GPU, and one Ollama gotcha

Google's Gemma 4 12B is the multimodal open model that fits a mid-range card — if you pull the right Ollama tag. The honest VRAM math, the cheapest GPU that runs it, and how to start.

Tensor · Sep 26 · 7 min