GLM API · OpenAI-compatible · one key, all models
The same GLM API, up to 50% off
Same Z.ai upstream, same quality — models priced at 60–90% of official plus the top-up bonus bring it as low as 50% of the official price. Change one line of base_url and your existing OpenAI SDK just works. Try it below, sign in when you are ready.
Best for general AI apps, agents, search, and high-volume API workloads
Z.ai official
$1.40
/ million output tokens
flatkey · effective price with top-up bonus
$0.56
/ million output tokens
↓ Top up $200, get $300 — stretch your token budget 1.5×
Playground (edit before sign-up)
model: glm-5.2temp 0.7 · 1024 tok
temperature 0.7max_tokens 1024top_p 1.0
▶ Sign in to runGoogle / GitHub one-click · no credit card to startEst. this run ≈ $0.003 (flatkey · official ≈ $0.006)
Pricing vs official
| Coverage | GLM 4.7 · GLM 5 Turbo · GLM 5.2 |
| Cache reads | up to 50% off |
* Illustrative pricing — see flatkey pricing page
Live model data from pricing API
glm-5.2
Z.ai
Input / 1M
$1.4
Output / 1M
$4.4
anthropic
Matched live models
glm-4.7
Z.ai
glm-5-turbo
Z.ai
glm-5.2
Z.ai
AI app backends
Agent workflows
Batch content generation
migrate.py — change one line
# Your existing OpenAI code: from openai import OpenAI client = OpenAI( base_url="https://router.flatkey.ai/v1", api_key="sk-flatkey-...", ) response = client.chat.completions.create( model="glm-5.2", messages=[{"role": "user", "content": "Hello"}], ) print(response.choices[0].message.content)
Does this use the same model id in my SDK?
Yes. Keep your SDK and switch base_url plus api_key.
Can I control usage before scaling?
Yes. Prepaid balance, usage analytics, and one invoice keep spend bounded.
Every top-up earns bonus credit
Pay to unlock · credited instantly · not a free-signup giveaway
Top up $10 get $3Starter / individual
Top up $200 get $100Team / high-volume