11 Programming Learning Services Compared [August 2026]: Choosing by Purpose, Price, End State and Language Support

Search for "best programming learning sites" and you get lists of thirty, mixing free services with schools that cost thousands of dollars. What the person choosing actually needs to know is which one to use, in what order, for their own goal.

This article compares 11 programming learning services, Japanese and international, organized by purpose: getting hired, finishing your first course, having something you want to build, rebuilding your fundamentals, and moving into infrastructure and AI. Mentor-led schools (a different category, at a very different price) are out of scope; everything here is a service you can study on independently for a few thousand yen a month at most. All prices include tax and reflect official figures as of August 2026.

The short version, by purpose

The 11 services at a glance

How to choose: four axes

1. Choose by what you will be able to do at the end

Even at a similar monthly price, the end states differ completely.

2. Get ahead of where people quit

Self-study stalls in a small number of predictable places.

3. Look at the yearly cost

Monthly figures are hard to compare, so here is a year of each.

4. Does it work in your language

Boot.dev, KodeKloud and DeepLearning.AI have no Japanese interface. Getting comfortable with technical English is an asset in itself, but making your first course an English one raises the odds of quitting. Build the base in your own language and cross over for the areas that require it — real-environment DevOps practice, AI implementation.

Recommendations by purpose

The services in detail

Here is each service by who it fits and how far it takes you. Even at the same monthly price, the roles do not overlap.

Progate | Finish your first course with no environment setup

Illustration-led slides for the concept, then an in-browser editor to write and run the code. 3.7 million users across 100+ countries. Its real value is that it removes the first wall, environment setup — the point where self-study most often ends.

Because the learning paths are laid out, you never wonder what comes next. Plus covers introductory through basic; Pro reaches work-level material such as Practical Backend Development, Practical Python and React Development. Command Line and Git lessons, Generative AI Basics and Claude Code Introduction (beta) all live in the same service.

ドットインストール | dotinstall: pick only the technology you need

576 lessons across 8,618 videos, watched three minutes at a time. The psychological cost of starting is very low, which makes it work in the gaps of a day. Running since 2011, it spans web production, game development, data analysis and AI/ML.

The caveat is that there is no in-browser runtime. You are expected to build the same thing in your own environment while watching, so "just watching" will not stick. On the other hand, lessons on local setup, Linux commands and Git make it effective when you want to clear up an environment question quickly.

paizaラーニング | paiza Learning: the exit connects to job listings

Three-minute videos alternate with exercises, and code runs in the browser. The free tier alone gives 928 courses and 1,209 exercises, so you can get a long way toward working knowledge at zero cost.

The distinguishing feature is the exit. A skill-check rank (847 problems) works directly as a job application credential, letting you demonstrate ability in code rather than on a résumé. The curriculum is itself built from 20M+ skill-check attempts and job data from 3,000+ companies.

ZEN Study | Environment through CS theory for ¥1,100 a month

Entirely original material written by working Dowango engineers, with coverage that stands out for the price. The programming track targets "from no experience to employable as a web engineer," and the same subscription also covers university-entrance preparation, web design and video creation.

The introductory course lists Linux, Git, GitHub, Docker and web security among its learning elements beside JavaScript — a practical choice. Beyond it lie large-scale web development in Scala and CS courses on data structures and algorithms, automata and networking. Self-study material is complemented by live classes and a Q&A community.

Techpit | Work backwards from what you want to build

A marketplace of tutorials by working engineers. Because you choose by the finished product — "build a matching app in Rails" — it addresses the stall that follows syntax. Add features afterwards and the result becomes the base of a portfolio.

You work in a local development environment; there is no in-browser runtime. If environment setup worries you, clear it first with a free service. INTLOOP Inc. acquired the business in November 2025.

Recursion | Break the plateau that follows "I can write code"

800+ hours built by former Meta engineers from US university CS curricula, at a stated 20% input, 80% output ratio — most of your time goes into assignments. There are 100+ backend assignments that avoid frameworks, plus a team development program.

It covers data structures and algorithms, OOP, databases, networking, operating systems and system design, entirely in Japanese. Its stated goal, "escaping copy-paste engineering," describes the reader exactly: most effective for people who finished the introductory stage and stopped progressing. Conversely, it is far too heavy if you have never touched HTML and CSS.

WebTerm | For people stuck on the environment, in the first two weeks

A learning Linux terminal that runs in the browser, implementing 100+ commands from pwd, ls and cd to grep, find, chmod, pipes and redirection, Vim and Git. Nothing to install, no account — the first command is one page load away.

Its 15 tutorials grade on the state of the filesystem after your command runs, so copying answers gets you nowhere and you end up typing until it sticks. Six tutorials cover Git troubleshooting alone, including undoing commits and resolving merge conflicts.

WebTerm Learn | Lock in the terminal and Git, in order

A 12-course, 129-lesson platform alternating illustrated slides with exercises: the terminal (introduction 20 / fundamentals 13 / advanced 25) and Git (introduction 23 / fundamentals 18), each stage followed by training for repetition. Every course is free; all it needs is a free account.

Where WebTerm is the practice environment itself, this is the curriculum that runs through it. The advanced terminal course reaches shell customization and SSH, taking you to the point of working on a server. Vim and the terminal multiplexer Herdr are also covered.

Boot.dev | Backend and operations along the same path

A challenge-based service from Ogdolo LLC. Python, Go, TypeScript and SQL are taught in the same curriculum as Linux, Git, Docker, Kubernetes, CI/CD and AWS, so the gap between writing code and deploying it rarely opens. More than 1.2 million learners.

The AI tutor "Boots" replies with Socratic questions rather than answers. Combined with a format that will not advance without working code, it is hard to stay passive in. The interface and content are English-only.

KodeKloud | Break infrastructure and fix it, in real labs

DevOps, cloud and AI learned through hands-on labs on real environments: 80+ courses, 880+ labs, and per-specialty environments for Kubernetes, Docker and Terraform that start instantly. Exam preparation for CKA/CKAD, AWS and Azure is included.

It does not teach language fundamentals, so it suits people past the syntax stage or those on the infrastructure and operations side. The AI learning path (19 courses, ~106 hours) includes Claude Code For Beginners (8h32m) and Cursor AI.

DeepLearning.AI | AI implementation, through free short courses

The AI learning platform founded by Andrew Ng, with 104 short courses of one to three hours, most co-produced with the companies behind the tools. That keeps the material current in a way books and generic online courses are not.

Short courses enroll for free. To rebuild machine learning and deep learning systematically, you move to the specializations and professional certificates delivered through Coursera. The prerequisite is being able to read code.

An order that does not break you

Assuming an hour on weekdays and three hours on weekends, here is a realistic progression.

Spending the first two weeks on the environment reduces friction at every later stage — and it costs nothing.

What to learn if you use AI coding agents

In 2026, coding agents such as Claude Code and Cursor entered everyday work. That did not remove the need to learn; it swapped out what the foundation consists of.

Agents write code, but reading the proposed diff, interpreting the error and recovering from a broken state remain human work — and all of it happens in the terminal. Progate now covers CLAUDE.md, skills and permissions in its Claude Code Introduction (beta); KodeKloud gives 8 hours 32 minutes to Claude Code For Beginners; DeepLearning.AI co-produces courses with Anthropic. The learning services themselves are being rebuilt around agents.

Which services to combine, and how, is covered in "CLI and programming study in the age of AI coding agents".

FAQ

Can I learn with free services alone?

Terminal and Git fundamentals are fully covered for free (WebTerm, WebTerm Learn). Language syntax goes a long way on paiza Learning's free tier (928 courses, 1,209 exercises). Paying starts to matter when you move into systematic CS fundamentals or real infrastructure practice.

Coding school or self-study?

Try a few thousand yen a month for one or two months first and see whether self-study actually turns over for you. For people who progress alone, a school's return on cost falls; for people who cannot keep going without support, human help is worth a great deal. Building that evidence first prevents the wrong choice.

Should I use several services at once?

Stacking services with different roles works well. "Environment = WebTerm (free)," "syntax = Progate," "building = Techpit" complement rather than duplicate each other. Subscribing to two services in the same role (Progate and dotinstall) adds little; pick one.

Any advice for not quitting?

Set the target small. Not "become an engineer" but "be able to commit and branch in Git by the end of this week" — a size you can verify in one or two weeks, so progress stays visible. Starting with something that runs entirely in the browser works for the same reason.

Conclusion

Choose a learning service by its end state and where you are currently stuck, not by price.

• Stuck on the environment → WebTerm / WebTerm Learn (free)
• Want to finish a first course → Progate / dotinstall
• Want a route into hiring → paiza Learning / ZEN Study
• Have something to build → Techpit
• Want a thicker foundation → Recursion / Boot.dev / KodeKloud / DeepLearning.AI

This article reflects information published on each company's official site as of August 2026. Pricing and course structure change; check the official sites for the latest details.