For technical leads, architects and product owners who must pick, justify and revisit LLM choices across OpenAI, Anthropic, Google, the hyperscalers and open-weight models.
Level
Intermediate
Study time
12 hours
Lessons
15 in 5 modules
Exam
20 questions
About this course
Model choice is now an ongoing engineering and commercial decision, not a one-off pick. New models ship monthly, the same model is sold through several channels with different terms, and published leaderboards rarely predict how a model will behave on your documents, your tools and your latency budget.
This course gives you a repeatable process. You will learn how the market is structured (frontier closed models, open-weight models, and the hyperscaler platforms that resell both), which capability dimensions actually separate models, how to read benchmarks critically, and how to build and run a task-specific evaluation. You will model cost per completed task, including prompt caching, batch discounts, reasoning tokens and the total cost of self-hosting. You will check data residency, retention and enterprise terms, read open-weight licenses for the clauses that matter, and design for portability and multi-model routing. The course ends with a weighted selection scorecard and a full worked example you can adapt. Specific prices and product facts are stated as of October 2026 or clearly labeled as illustrative.
What you will be able to do
Map a use case onto the model market and identify which channels (direct API, hyperscaler platform, self-hosted open weights) are viable for it
Critique a published benchmark claim and explain why it may not transfer to a specific workload
Build a stratified, task-specific evaluation set with gold answers, a grading rubric and a holdout split
Calculate cost per completed task, including caching, batch, reasoning tokens and retries, and estimate a self-hosting break-even point
Evaluate data handling, residency and enterprise terms and open-weight license clauses against organizational requirements
Design a portability layer and a multi-model routing and fallback policy that limit lock-in
Apply a weighted selection scorecard with must-have gates and produce a defensible decision record
Syllabus
5 modules · 15 lessons · about 12 hours
1
Module 1: The Model Market and What Separates Models
How the 2026 market is structured, what changes when you buy the same model through different channels, and the capability dimensions that matter.3 lessons
The 2026 Market Map: Frontier Closed, Open-Weight and the Platforms Between
Direct API vs Hyperscaler Platform: Buying the Same Model Through Different Doors
Capability Dimensions That Actually Separate Models
2
Module 2: Evidence: Benchmarks and Your Own Evals
Reading public benchmarks with skepticism and building the task-specific evaluation that should drive the decision.3 lessons
Reading Benchmarks Critically
Building a Task-Specific Eval Set
Running a Fair Head-to-Head Comparison
3
Module 3: Cost Modeling
From list price per token to cost per completed task, the discount levers, and when self-hosting open weights pays off.3 lessons
Token Economics and Cost per Completed Task
Caching, Batch, Service Tiers and Reasoning Tokens
Self-Hosting TCO for Open-Weight Models
4
Module 4: Terms, Licenses and Lock-In
Data handling and residency, open-weight license clauses, and engineering for portability.3 lessons
Data Handling, Residency and Enterprise Terms
Open-Weight Licenses: What You Can Actually Do
Lock-In and Portability
5
Module 5: Deciding and Operating
Routing across models, a repeatable selection scorecard, and an end-to-end worked selection.3 lessons
Multi-Model Routing and Fallback
A Repeatable Selection Scorecard
Worked Example: Selecting Models for a Claims-Intake Assistant
Exam and certificate
The exam unlocks once you have completed every lesson. It draws 20 questions from a larger bank, spread across the modules, with the answer options shuffled for each attempt. You have 40 minutes and need 80% to pass. After a failed attempt you can try again after 24 hours, up to 3 attempts in total. Feedback is given per module, never as the answer key.
Passing earns a certificate of completion and assessment from the Wirewalk Institute of AI with a unique ID that anyone can check at institute.wirewalk.com/verify. It is not an accredited qualification or a vendor certification. How certification works.
The Wirewalk Institute of AI is independent. This course teaches publicly documented products and is not affiliated with or endorsed by OpenAI, Anthropic, Google (Gemini, Vertex AI, Gemma), Microsoft (Azure OpenAI, Microsoft Foundry), Amazon (Bedrock), Meta (Llama), Mistral AI, DeepSeek, or Alibaba (Qwen). Product and company names are trademarks of their respective owners.
A hands-on hour with ChatGPT, Claude and Gemini, told through one small business racing a deadline. Set up an account with a privacy setting you chose, decide what never goes into a personal account, write prompts that produce usable drafts, use files and web search, catch confident mistakes before they go out, and pick the assistant that fits how you work.
A practical, current tour of ChatGPT as a professional tool: choosing a plan and model, prompting well, analyzing files, researching with citations, using images and voice, organizing work in projects and memory, connecting apps and plugins, delegating multi-step work to ChatGPT Work, and knowing which admin and privacy controls matter. Ends with clear guidance on moving a workflow to the OpenAI API.
Learn Claude by solving one firm's real problem: getting through renewal season with too few analysts and without putting client data in the wrong place. Each module compares the approaches professionals actually choose between, shows a realistic failure and its fix, and ends with a Decision Lab where you make the call under pressure.