Model
.ai
tracking open ai models
Home
/
Text / LLM Models

Text / LLM AI models

Text / LLM groups the open large language models in the catalog: Llama, Qwen, Gemma, Mistral, DeepSeek, Phi and the fine-tunes built on them, from 1 billion parameter models that run on a laptop to frontier-size mixtures of experts. Each entry lists the parameter count, context length, licence and where to run it; the models Graydient hosts carry a Runs badge and a token for the PirateDiffusion chat bot.

Hottest Text / LLM models

Text / LLM guides

Cloud services
Flat fee vs tokens: why unlimited plans like Graydient.ai are the best value for AI creators in 2026
A ranked cost comparison, with prices checked on 8 October 2026, of flat-fee unlimited plans like Graydient.ai against token and credit pricing from Midjourney, Leonardo, fal.ai, Replicate, Runway and Kling: what 1,000 images and 1,000 videos a month really cost on each, and why one fixed fee for unlimited images, video, audio, Grok LLM chat and web apps is the best value for creators who iterate.
Apps
How to Install Ollama and Run a Local AI (Windows, Mac, Linux)
You can run a ChatGPT-style AI on your own PC or Mac for free, offline and with no per-token fees. This guide shows how to install Ollama, which model to download for 8, 16, 24 or 32 GB of memory, how to chat with it, and what to do when it is slow or runs out of memory.
Apps
Ollama vs LM Studio (2026): which local AI app to use, and which top models each one runs
Ollama and LM Studio are the two free ways to run a ChatGPT-style AI on your own computer, and both run the same open models. Ollama is open source, light and built to be the engine behind other apps; LM Studio is a polished desktop app with a model browser and the faster engine on Apple silicon Macs. This comparison covers speed, Mac, NVIDIA and AMD support, the local server, coding agents and privacy, and which of this week's top-ranked open models each one can run on your hardware.
Models
Best local LLM by use case (October 2026): the right open model for every job
There is no single best local LLM: the right open model depends on the job and on your memory. Using this week's text model rankings, this guide picks one model per use case (everyday chat, coding, coding agents, reasoning, reading images, long documents, translation, creative writing, laptops, searching your own files and voice), with the memory each one needs, one alternative, and the commands to run them. It also explains why the #1 ranked open models need a server, and which of their smaller siblings fit at home.
Models
Civitai alternatives: where to find and download AI models in 2026
Civitai is still where most LoRAs are born, and it is also a single company that has removed whole categories of models, changed its payment rules and gone dark for days under pressure from payment processors. If your workflow depends on a file only Civitai hosts, you have a problem waiting to happen. Captain's map of everywhere else the open-weight world keeps its models in 2026: Hugging Face, the model-page mirrors, torrents and The Hugging Bay, AI Horde, community archives, the creators' own channels, and how to use this site to find the same model in several places at once. Plus how to keep a local archive so no takedown ever stops your work.
Errors & fixes
How much VRAM do I actually need for local AI?
The honest answer, by model rather than by budget: 8 GB runs 8B chat models and SDXL, 12 to 16 GB runs 14B models, Flux fp8 and short video, 24 GB is the sweet spot for 30B-class LLMs and FLUX.2 or Qwen Image, 48 GB and up is 70B territory, and 64 GB of system RAM is not a 64 GB GPU. With the arithmetic for weights, context and headroom, a memory tier table from 8 GB to 512 GB, and Kai's video on every way to reach 128 GB of VRAM.
All guides

Other model families