Your magic wand for everyday numbers and AI
Clear guides to the calculations of daily life, with free calculators — and a way into AI: tutorials, the latest models, the companies behind them, and our own local AI Lab.
Everything in one place
The numbers of daily life, explained
Money
Loans, interest, percentages, rent, cheques and gold.
Home & energy
Electricity use, paint and tiles.
Travel
Fuel and the cost of a trip.
Time & dates
The Iranian and Gregorian calendars, and Iran's time zone.
Study & numbers
Averages, medians and weighted grades.
Digital
Data sizes and internet speed.
Worked examples you can follow
How loan instalments are calculated, with a worked example
The standard formula behind equal monthly loan payments, a worked example for a 24-month loan, and how much of each payment is interest.
Percentages in everyday life: discounts, increases and points
How to work out discounts and increases, why two discounts don't simply add up, how to reverse a percentage, and percent versus percentage points.
Converting rent and deposit (rahn and ejareh)
How the Iranian rent-and-deposit system works, the common conversion convention between deposit and monthly rent, and a worked example.
Understanding your electricity use: watts and kilowatt-hours
How to turn an appliance's watts into kilowatt-hours, which appliances use the most, a worked example, and simple ways to lower the bill.
Solar Hijri and Gregorian dates: how the two calendars line up
How the Iranian solar calendar relates to the Gregorian calendar — year offsets, month lengths, leap years — and Iran's time zone.
Fuel and trip cost: working out what a journey costs
How to calculate fuel use and fuel cost for a trip, convert between L/100 km and km per litre, and plan a realistic travel budget.
Do the sums for your own numbers
Loan calculator
Monthly payment, total interest and loan comparison.
Bank interest
Simple interest, compound interest and deposits.
Rent and deposit
Convert deposit to rent and rent to deposit.
Cheque and note discount
Present value of cheques and promissory notes, discount and interest.
Currency and gold
Convert currencies and work out gold values.
Fuel and trip cost
Trip cost from your car's consumption and the fuel price.
AI tutorials
Prompt engineering masterclass
Practical techniques for getting better results from any large language model: clear tasks, context, examples, formats, step-by-step work and checking.
Image generation: writing prompts that work
How AI image generators such as DALL-E, Midjourney, Stable Diffusion and FLUX work, how to write a good image prompt, and how to use the results responsibly.
AI coding assistants: Copilot, Cursor, Claude Code and more
What AI coding assistants do — completion, chat and agents — how GitHub Copilot, Cursor and Claude Code differ, local options, and how to use them safely.
Video and audio AI: from speech to music and moving images
What AI can do with sound and video — speech recognition, text-to-speech, music and text-to-video — the main tools, and the ethics of cloning voices.
AI for business: automating work without losing control
Where AI helps a business — drafting, summarising, support, search over documents — how to run a small pilot, keep people in the loop, and protect data.
Local AI setup: Ollama, LM Studio, llama.cpp and vLLM
Tools for running open AI models on your own hardware — Ollama, LM Studio, llama.cpp and vLLM — what each is for, quantisation, and the memory you need.
AI models
Large language models
GPT-4o, Claude, Gemini, Llama, Mistral and Mixtral — chat, writing, reasoning and analysis.
Image and vision
DALL-E 3, Midjourney, Stable Diffusion / SDXL, FLUX.1, CLIP — creating images, and understanding them.
Code
Codex, StarCoder, Code Llama, DeepSeek Coder, Qwen2.5-Coder — writing, completing and explaining code.
Audio and speech
Whisper, XTTS, MMS, Bark, ElevenLabs — transcription, voices and music.
Video
Sora, Runway Gen-3, Veo, Kling, Pika — text-to-video and video editing.
Open models to run yourself
Llama, Mistral, Phi, Qwen, Gemma — download once, run privately on your own hardware.
AI companies
OpenAI
Creators of the GPT models, DALL-E, Whisper and ChatGPT, whose launch in November 2022 brought AI chat to hundreds of millions of people.
Google DeepMind
DeepMind was acquired by Google in 2014; its AlphaGo beat world champion Lee Sedol in 2016, and AlphaFold predicted protein structures — work recognised with the 2024 Nobel Prize in Chemistry.
Anthropic
Founded by former OpenAI researchers including Dario and Daniela Amodei.
Meta AI
Meta's AI research lab created PyTorch, the widely used deep-learning framework, and releases open models such as Llama, SAM and NLLB for anyone to run.
Mistral AI
A European AI company building efficient, open-weight language models, from Mistral 7B to the mixture-of-experts Mixtral, alongside commercial models and the Le Chat assistant.
More companies
xAI, Stability AI, Runway, Midjourney, Cohere, Hugging Face and the growing AI ecosystem.
AI Lab — ai.zibadis.ir
56 local models, 17 categories
AI Lab is our hands-on AI application: interact with models, run experiments and explore AI directly — text and code, vision, speech, OCR, translation, 3D and more, all on ZIBADIS servers with no external API.
Privacy
Your data never leaves ZIBADIS's servers — fully local processing with no external service calls.
No per-token costs
Local models don't charge per word, so experimenting costs nothing extra.
Speed & reliability
No dependence on international internet — answers keep coming even when networks are disrupted.
Recent articles
Getting started with Claude Code
A beginner's guide to Claude Code, Anthropic's AI coding assistant that works in the terminal: installation, first steps, and working with it safely.
GPT-4o vs Claude Opus vs Gemini Ultra
How to compare large language models from OpenAI, Anthropic, Google and others for your own work: tasks, context, cost, privacy and testing.
Inside Anthropic's safety-first approach
How Anthropic approaches AI safety: Constitutional AI, its Responsible Scaling Policy and AI Safety Levels, and research into how models work inside.