AI Provider Setup · BYOK
200K context with optional thinking mode — strong tool use at lower price than Opus.
Z.AI's GLM family (GLM-4.5 and 4.6) offers 200K context windows, an optional extended-thinking mode for harder problems, and competitive tool-use performance — usually at a lower price than Claude Opus. Good middle-ground if you want reasoning power without the top-tier per-token cost.
ZAI_API_KEYSign up at z.ai. You'll get a small amount of trial credit; production use requires adding a payment method.
Open Billing in the dashboard and add a card. GLM keys do not work without an active billing method once trial credit is exhausted.
Navigate to z.ai/manage-apikey/apikey-list. This is the only place you can view or create keys.
Click Create, name it Admaxxer, and submit. The full key is shown once — copy it before closing the dialog.
Go to Settings then AI Providers, find the GLM row, paste the key, and click Connect.
glm-4.6 is our recommended pick — strong tool use at a competitive price. Turn on thinking mode for hard reasoning tasks; leave it off for speed.
Re-check the key was copied in full (includes the dot), billing is active, and your account region is consistent with the key.
Once you've pasted your key into /settings/ai-providers, hit Test. Admaxxer makes a single no-cost call against the provider's /v1/models (or equivalent) endpoint and reports back:
These are the pinned model IDs Admaxxer tests against. The model picker in chat will show these plus any live-catalog entries the provider returns.
| Model ID | Display name | Context | Tools | Best for |
|---|---|---|---|---|
glm-4.6 |
GLM-4.6 (recommended) | 200K | Yes | Best balance of reasoning, tool use, and price in the GLM family. |
glm-4.7 |
GLM-4.7 | 200K | Yes | Newer iteration with improved reasoning. Availability varies by region. |
glm-4.7-flash |
GLM-4.7 Flash | 200K | Yes | Faster, cheaper 4.7 variant. Good for high-throughput tool-use loops. |
glm-4.5 |
GLM-4.5 | 200K | Yes | Stable previous-gen flagship. Use if 4.6 is unavailable in your region. |
glm-4.5-air |
GLM-4.5 Air | 200K | Yes | Lighter Air variant of 4.5 — cheapest GLM tier that still supports tools. |
glm-4.5-flash |
GLM-4.5 Flash | 200K | Yes | Very fast 4.5. Good for classification and routing at volume. |
When it happens: On Connect if the key is wrong or was truncated on copy.
Fix: Regenerate at z.ai/manage-apikey/apikey-list. Make sure you copy both halves (before and after the dot).
When it happens: When trial credit is exhausted and no card is on file.
Fix: Add a payment method in z.ai billing.
When it happens: Under sustained parallel load.
Fix: GLM rate-limits per-tier. Upgrade tier in billing or slow down request rate.
When it happens: If your Admaxxer server region is blocked from the model you picked.
Fix: Pick a widely-available model like glm-4.6 or switch to GLM via OpenRouter for broader region coverage.
When it happens: During peak hours in China-hosted regions.
Fix: Retry with backoff. For critical flows, fail over via OpenRouter.