The full Hermes + GLM-5.2 setup lives inside the AI Profit Boardroom
Join
The New Hire · new framework

The Goldie New Hire™ + GLM-5.2 on Hermes.

New — GLM-5.2 now runs as a Hermes agent

Hermes runs a crew of AI agents for you — each one a different model you can swap in. I just added a new hire to the bench: GLM-5.2, the frontier coder, on a flat Coding Plan. It builds straight into my Hermes workspace while the rest of the crew does their jobs. Here's how it joins the team.

A 3D game GLM-5.2 built as a Hermes agent — a dragon over a mountain vista
1new agent on your crew
1Mtoken context
flatCoding Plan · no token meter
8things it built into the workspace
I.5 ────── watch the new hire

The promo my crew made — about the new hire.

I gave the Hermes crew one job: make a video about GLM-5.2. It researched the release, wrote the script in my voice, put it in my face with my cloned voice, edited it, and a judge graded it. Press play.

↑ researched, scripted, voiced, edited + judged by the GLM-5.2 Hermes crew

What the new hire built into Hermes — open everything ↓
play · gameThe Dragon Realm → play · gameNordic Crypt → play · appGLM OS desktop → play · gameNeon Blaster → the agentGLM 5.2 in Agent OS → the crewHermes workspace →
II ────── what the new hire shipped

I added it to the crew. Then it got to work.

Every tile is a real build GLM-5.2 wrote as a Hermes agent. They play right here on the page as you scroll — or hit ⤢ to open one full-screen.

⤢ full
The Dragon Realm
rebuilt the Kimi Skyrim game
⤢ full
Nordic Crypt
rebuilt the Kimi crypt game
⤢ full
GLM OS
rebuilt the Kimi web desktop
⤢ full
Neon Blaster
a juicy neon space shooter
⤢ full
Synthwave Drive
a neon sunset grid road
⤢ full
Particle Forge
a glowing particle sandbox
✦
my story · why this matters

I was you. Then I gave my Hermes crew a new hire.

Before

My coding agent was one model, locked in.

When a better one dropped, switching meant a new tool and a new setup.

My agents didn't share anything — different tabs, different memory, nothing saved in one place.

And the good coder charged me per token, so I rationed it.

Then I added GLM-5.2 to Hermes as its own agent.

After

Now GLM-5.2 sits on my crew next to my other agents.

I pick it from a switcher — same chat, same workspace, same board.

It builds straight into my Hermes workspace, where everything it makes is saved and playable.

And it runs on a flat plan, so I never ration it.

You can have this too. Same crew. Same path.

III ────── the receipts

Real people. Real wins. Inside the Boardroom right now.

Here's what's happening for the members already running this stack — agency owners, ecom founders, course creators, solo operators. Different businesses. Same engine room.

3,200+ Founders inside AIPB
38 Countries · live members
$100k+/mo AIPB MRR

What's already happening for members

Real member win
Real member · ranking + traffic win from the same Agent OS stack
 Real member win
Real member · output up without hiring
 Real member win
Real member · built their first agent the same day
 See all the wins (158-page doc) →
Before you scroll on —

Commit to transitioning today. Not tomorrow.

You've seen the builds. You've seen the proof.

The next few minutes show exactly how GLM-5.2 joins your Hermes crew.

So here's the deal.

Promise yourself one thing right now. You finish this guide, and you add one new agent to your crew before you sleep tonight. Just one. Because the moment your agents share one workspace and you can swap the best model in, your whole build process changes.

The people sitting still are stuck on one locked-in tool. The people who build a crew today are the ones who slot in every new model the day it drops.

Be one of those people.

Commit to the transition. Take action today. This changes how your agents work together.

✦
IV ────── the framework

The Goldie New Hire™.

Treat every new model like a new hire on your Hermes crew. Five plain steps to put it to work — and to swap in the next one the day it drops.

i.

The Hire

You add the model as its own Hermes profile. GLM-5.2 becomes an agent you can pick — no new app, no new window.

ii.

The Pay

You put it on a flat Coding Plan, not a per-token meter. The new hire works all day for one predictable price.

iii.

The Desk

You give it your Hermes workspace. Everything it builds lands in one place — saved, listed, playable — not lost in a chat.

iv.

The Crew

You sit it next to your other agents. Kimi, N2, GLM-5.2 — pick the right one for the job from one switcher, on one board.

v.

The Swap

You never get locked in. When a better model drops, you hire it the same way — one profile — and the crew gets stronger.

V ────── old way vs new way

One locked-in coder. Or a crew you pick from.

The old way
one tool, locked
  • One coding tool, one model, hard to change
  • A new model means a whole new setup
  • Agents in different tabs, no shared memory
  • Nothing saved in one workspace
  • Pay per token, so you ration the good one
  • You're always a version behind
The new way
a Hermes crew
  • Every model is an agent you pick from a switcher
  • Add a new one as a profile in minutes
  • One chat, one workspace, one kanban
  • Every build saved + playable in the workspace
  • GLM-5.2 on a flat plan — no rationing
  • Slot in each new model the day it drops
Thinking it?"I already have a coding agent. Why add another?"

Because you stop betting on one model. A crew lets you pick the best one per job — and GLM-5.2 is a frontier coder you don't have to meter.

Adding it doesn't replace what you have. It joins it.

✦
VI ────── the hire

Add GLM-5.2 to Hermes in two steps.

Hermes already speaks z.ai. You make a profile, point it at the Coding Plan endpoint, and hand it your key.

profile: glm-5-2 provider: zai model: glm-5.2 base_url: https://api.z.ai/api/coding/paas/v4

Then add the key to that profile, and it's hired. Pick "GLM 5.2" from your agent switcher and start chatting.

Thinking it?"That looks like a config file. I'm not technical."

Inside Agent OS it's a dropdown and a paste box — no file editing.

And members who'd never opened a terminal already run agents like this.

VII ────── the desk

It builds into your workspace, not a chat.

Ask the new hire to build something. The finished file lands in your Hermes workspace — the "GLM 5.2 ✦" bucket — saved and playable.

No copy-pasting code out of a chat window. No lost work. You open the bucket and click the thing it made.

GLM OS — a web desktop GLM-5.2 built into the Hermes workspace
Thinking it?"Will this clutter up my setup?"

No — the new agent gets its own bucket, separate from your other agents' work.

One crew, but everyone has their own desk.

VIII ────── the crew

Put it on the board with the rest.

GLM-5.2 isn't a loner. On the Hermes kanban you can hand it a card like any other agent.

So a job can flow across the crew — one agent writes, GLM-5.2 builds, another reviews — all on one board you watch.

"The point of a crew isn't one super-agent. It's the right agent for each job, in one place."

— how Hermes runs the team
IX ────── straight from Z.ai

The official sources. Read it and run it yourself.

The model details come from Z.ai's own announcement and docs. Here's where to read it and grab a key:

Official Z.ai sources + resources ↓
the announcementZ.ai — GLM-5.2 → docs · how to switchSwitch to GLM-5.2 → docs · homeZ.ai Dev Docs → get startedThe Coding Plan →

"GLM-5.2 is now available to all GLM Coding Plan users… As our new flagship model, GLM-5.2 delivers powerful coding capabilities, usable 1M-context support, and continued strengths in long-horizon tasks."

— Z.ai (@Zai_org), 13 June 2026

✦
X ────── the shortcut

Want the whole Hermes crew, already wired?

If you want GLM-5.2 sitting on a crew with Claude, Kimi and your other agents — sharing one memory, one workspace, one board — get the Agent Operating System inside the AI Profit Boardroom.

It's the full system I built. Your agents know your business, so the new hire builds in your style from the first prompt.

Get the Agent OS →
Inside the AI Profit Boardroom · aiprofitboardroom.com
link in the description ↓
XI ────── what's really stopping you

Three beliefs holding you back.

"I should pick one AI and stick with it."
The models change every few weeks. A crew lets you swap the best one in without rebuilding your whole setup.
"Running multiple agents sounds complicated."
It's one dashboard. You pick an agent from a switcher. Hermes handles the rest.
"I'll set up a proper agent crew later."
"Later" is how you stay locked to one tool while everyone else slots in each new model on day one.
Don't take my word for it

158 pages of members who already broke through these exact beliefs. Real businesses, real wins, documented.

Read the 158-page testimonials doc →
✦
XII ────── how to run it

Hire GLM-5.2 onto your crew in 5 minutes.

1

Get on the Z.ai GLM Coding Plan and grab your key.

2

In Hermes, make a profile called glm-5-2.

3

Set provider zai, model glm-5.2, base URL https://api.z.ai/api/coding/paas/v4.

4

Add your key to that profile.

5

Pick "GLM 5.2" from the agent switcher and say hello.

6

Ask it to build something — it lands in the GLM 5.2 ✦ workspace bucket.

7

Hand it a card on the Hermes kanban for bigger jobs.

8

When the next model drops, hire it the same way.

XIII ────── recap

What you just got.

i.
You got a new hire.

GLM-5.2 is an agent on your crew now.

ii.
You stopped paying per token.

A frontier coder on a flat plan.

iii.
You stopped losing work.

Every build saved in the workspace.

iv.
You stopped tab-hopping.

One chat, one workspace, one board.

v.
You stopped getting locked in.

Swap the best model in, any time.

vi.
You got a real crew.

The right agent for each job, in one place.

Every new model. Just another hire.

Build your crew around it.

If you want GLM-5.2 working alongside all your agents — not in its own lonely tab — grab the Agent Operating System inside the AI Profit Boardroom.

It turns Claude, Hermes and your agents into one system with shared memory and one dashboard. The new hire slots right in, already knowing your business.

Get the Agent OS →
Inside the AI Profit Boardroom · aiprofitboardroom.com
I'll see you in the next one.