Academy
Buy AI Academy$399$199/yr
Buy AI Academy — $399$199/yr
Coding on Open Models
PractitionerEngineer
Run the agent harness you already know on open backends (GLM, Kimi, DeepSeek, Qwen), measure the gap yourself, and route by task.
6 chapters ~1.5 hr
Earn a Certificate of Completion
GLM & Z.ai: The Complete Guide
PractitionerNon-tech
Everything GLM: the 5.x line, the Coding Plan, the two-line Claude Code switch, ZCode, how it stacks up against Claude and GPT, and what an enterprise should check before betting on it.
7 chapters ~2 hr
Earn a Certificate of Completion
Kimi: The Complete Guide
PractitionerNon-tech
Everything Kimi: who Moonshot is, the K-series through K3, Kimi Code hands-on, and the honest answer to is it comparable to Claude Code.
7 chapters ~2 hr
Earn a Certificate of Completion
Open-Source AI Models: Literacy & Landscape
FoundationsNon-tech
Read the open-weight AI model market like a strategist — who leads, what the licenses really permit, how to access them, and who to bet on.
7 chapters ~2 hr
Earn a Certificate of Completion
The Open-Source Way
PractitionerNon-tech
The adoption playbook: how a company actually moves from interesting to open models in production, covering sovereignty, cost, pilots, and exit ramps from lock-in.
7 chapters ~2 hr
Earn a Certificate of Completion
Serving Open Models in Production
AdvancedEngineer
Self-hosting as an engineering discipline: vLLM, GPU math, OpenAI-compatible gateways, observability, the honest cost model, and hardening a shared GPU across tenants.
7 chapters ~2 hr
Earn a Certificate of Completion
6 chapters
Explore the full Coding on Open Models path
Coding on Open Models: Start Here
Five chapters on wiring, measuring, and routing open coding models into a real engineering team's stack, in the order that keeps you from redoing work.
11 min
Wiring Open Backends
Redirect Claude Code's own connection settings to GLM, Kimi, and other Anthropic-compatible backends, and know exactly what changes when you do.
16 min
Benchmark It Yourself
Build a small personal eval harness on your own repo, because the public leaderboard is answering a different question than the one you're asking.
15 min
Routing by Task
Send planning to a frontier model, implementation to mid-tier, and quick edits to whatever's cheapest, using your own eval numbers instead of vibes to draw the lines.
15 min
The Failure Modes
Where open coding models still break under real workloads, and the checkpoints, gates, and task splits that catch it before it costs you a day.
15 min
Team Rollout of a Hybrid Stack
Seats, spend caps, and policy that make an open-model addition to the stack survive contact with a whole engineering team, not just your own terminal.
18 min
Certificate
Certificate of mastery
Complete all chapters to earn your certificate
7 chapters
Explore the full GLM & Z.ai: The Complete Guide path
GLM & Z.ai: Start Here
Six chapters, one fast-moving model family from a different vendor. The map before you pick a chapter, and the order that actually makes sense.
11 min
The GLM-5 Line Explained
Three releases in four months, one very deliberate context-window jump: what changed at each step, and why it matters before you adopt anything downstream.
18 min
The GLM Coding Plan
Three tiers built on one shared credit system, what each actually buys, and how to size the right one before you commit.
17 min
GLM Inside Claude Code
Pointing an existing Claude Code setup at the GLM Coding Plan takes no new tool to learn, only redirected traffic, so here's exactly what changes and what quietly stays put.
16 min
ZCode & the Tool Ecosystem
Z.ai ships its own desktop agent built around GLM, and a long list of tools a team may already use can connect to the same models too, here's how to tell which path actually fits.
18 min
GLM vs. Claude vs. GPT
Three models are competing for the same coding work today, and knowing where each one actually pulls ahead beats trusting whichever leaderboard screenshot crossed your feed last.
19 min
GLM for the Enterprise
Running GLM at a company raises two separate questions, where the model runs and what happens to the data, and both deserve a real answer before either becomes the default.
16 min
Certificate
Certificate of mastery
Complete all chapters to earn your certificate
7 chapters
Explore the full Kimi: The Complete Guide path
Kimi: Start Here
Six chapters, one fast-moving model family. The map before you pick a chapter, and the order that actually makes sense.
11 min
The K-Series Explained
K2 through K3, what changed at each release, and which weights and licenses are actually open to you today.
18 min
Kimi Code Essentials
Install it, get through a safe first session, and pick the plan that actually matches how your team will use it.
17 min
Kimi Code Pro Flows
When one agent isn't enough, and how to tell whether it actually is.
19 min
Kimi vs. Claude Code vs. Codex
The benchmarks, the real pricing, and the day-to-day feel, read honestly, then turned into one routing table you can actually reuse.
18 min
Kimi Inside Your Existing Tools
You don't have to switch tools to try Kimi. Here's what bringing it in as a backend actually gets you, and what it doesn't.
17 min
Running Kimi Yourself
The weights are downloadable today. What that actually demands, and when the honest answer is still to use the API.
18 min
Certificate
Certificate of mastery
Complete all chapters to earn your certificate
7 chapters
Explore the full Open-Source AI Models: Literacy & Landscape path
Open-Source AI Models: Start Here
A 12-minute orientation to reading the open-weight model market like a strategist — what each chapter gives you, how they connect, and where to begin.
12 min
Open Source vs Open Weight
The word "open" hides two very different deals — and the one that decides what you can legally ship is the license, not the label.
18 min
The 2026 Open-Model Leaders
A working map of who leads the open-weight field right now — and the bigger story of where the lead has moved.
18 min
Reading Open-Model Benchmarks
The leaderboards that rank open models, the self-reported numbers that flatter them, and how to read an open-vs-open comparison without being fooled.
18 min
What Open Models Cost
Two cost worlds — renting by the token and running it yourself — and why "free weights" lands you in the more expensive one more often than people expect.
18 min
How to Access Open Models
The providers that serve open models for you, the aggregator that hands you all of them through one door, and how to pick — by price, speed, and where your data may go.
18 min
Who to Bet On
A durability framework for picking an open-model vendor that will still be strong and supported next year — and how to hedge so no single bet can sink you.
18 min
Certificate
Certificate of mastery
Complete all chapters to earn your certificate
7 chapters
Explore the full The Open-Source Way path
The Open-Source Way: Start Here
The 2026 shift toward open weights within striking distance of frontier at a fraction of the price, and the six-decision playbook for moving from curiosity to production.
9 min
Three Ways to Consume Open Models
Hosted API, aggregator, or self-hosted: the four-question matrix that tells you which door fits your workload.
18 min
Sovereignty & Compliance
A due diligence checklist for data residency, license terms, and origin risk before an open model goes into production.
16 min
The Pilot That Proves It
A two-week bake-off template that turns a hunch about an open model into a verdict a budget owner can act on.
17 min
TCO, Honestly
A worksheet for building an honest cost comparison between hosted APIs and self-hosting, so the verdict comes from your own workload.
16 min
The Hybrid Stack
A routing matrix for deciding which work goes to your frontier model and which goes to an open one, and the discipline that keeps the split honest.
16 min
Exit Ramps & Vendor Durability
Portability tests for swapping the model behind your stack, and a who-to-bet-on framework for picking an open vendor built to last.
15 min
Certificate
Certificate of mastery
Complete all chapters to earn your certificate
7 chapters
Explore the full Serving Open Models in Production path
Serving Open Models: Start Here
Six chapters that take an open-weight checkpoint from a container that answers one request to a deployment that survives real traffic, a real budget, and more than one tenant.
10 min
vLLM & the Serving Stack
Why serving a model to many users at once is a different engineering problem than running one on your own machine.
17 min
GPU Sizing & Quantization at Scale
Turning weight count and context length into a real VRAM budget, instead of guessing and watching a container crash on startup.
17 min
The OpenAI-Compatible Gateway
One stable address in front of two correctly sized backends, with routing, failover, and health checks doing the work no client should have to.
16 min
Observability & SLOs for Inference
Turning a vague sense that requests feel slow into a percentile target you can actually alert on, and tracing which hop is responsible when it is breached.
17 min
The Cost Model, Honestly
Why the GPU hourly rate is rarely the number that actually sets cost per token, and what a tight latency target quietly costs in reserved headroom.
17 min
Hardening & Multi-Tenancy
Turning named-but-unconfigured virtual keys into an actual isolation guarantee, so one tenant's burst can never quietly spend another tenant's headroom.
17 min
Certificate
Certificate of mastery
Complete all chapters to earn your certificate