Models & LLMs
Large language models: prices per million tokens, benchmark scores, running models locally and the labs that build them.
7 articles
-
Mistral AI: models, pricing and agents
Mistral AI offers open-weight and closed models, an API and agent tools. Model lines, price per million tokens and typical use cases compared in one place.
-
LLM benchmarks: how to read model rankings
An LLM benchmark measures a specific skill, not intelligence. Learn to read model rankings by checking saturation, margins of error, and the cost of each gain.
-
Running a local LLM: hardware, software, and limitations
A local LLM runs on your computer using a downloaded model. Learn what hardware and software you need, and understand the limits before buying equipment.
-
Open-source or closed models for your AI agents?
Open models give you more control; closed models simplify operations. Compare licensing, total costs, and task reliability to choose a model for your AI agent.
-
The cheapest LLM for each use case
The cheapest LLM depends on how much you read, generate, and correct. Compare prices by use case and calculate what you pay for each successfully completed task.
-
Which GPU do you need to run an LLM?
Choose a GPU for a local LLM based on memory needs first. Check graphics cards, software support, and power requirements before spending money on new hardware.
-
Ollama step by step: install and run a model locally
Install Ollama on Windows, macOS, or Linux, download a model, and start your first local chat with clear commands and checks to confirm everything is working.