GLM 5.2 API — retired (upgraded to GLM 5.3)
GLM 5.2 was Zhipu AI’s (Z.ai) frontier open-source coding and reasoning model. It was served in the Frontier Pool until August 30, 2026, when the pool’s GLM slot was upgraded in place to GLM 5.3 — the same base model with extended post-training, at the same flat price. Frontier subscriptions were not interrupted and nothing changed on existing keys.
What this means
Section titled “What this means”- Requests with
model: "glm-5.2"are served by GLM 5.3 until September 30, 2026. After that date the id returns an invalid-model error. - Frontier Pool subscriptions are unchanged: GLM 5.3 and MiniMax M3 remain served with no token caps.
GET /v1/modelsis always the authoritative live list.
Migrate in one line
Section titled “Migrate in one line”Change the model id — the endpoint, key, and subscription stay the same:
from openai import OpenAI
client = OpenAI( base_url="https://api.cheapestinference.com/v1", api_key="sk-..." # your subscriber key)
response = client.chat.completions.create( model="glm-5.3", # was "glm-5.2" messages=[{"role": "user", "content": "Write a unit test for..."}],)curl https://api.cheapestinference.com/v1/chat/completions \ -H "Authorization: Bearer YOUR_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model": "glm-5.3", "messages": [{"role": "user", "content": "Hello"}]}'Coding agents: set the model to glm-5.3 in your tool’s config — see GLM with Claude Code, Cline / Roo Code and OpenCode, Roo & Aider.
GLM 5.2 in brief (historical)
Section titled “GLM 5.2 in brief (historical)”| Model ID | glm-5.2 (retired — use glm-5.3) |
| Served | Frontier Pool, until 2026-08-30 |
| Context window | 198K tokens |
| Open weights | Yes — zai-org/GLM-5.2, MIT license |
| Successor | GLM 5.3 — same 753B base, scaled post-training, open-weighted under the GLM-5.3 License |
Common questions
Section titled “Common questions”Is there still a GLM 5.2 API?
Not as a separate model. Requests for glm-5.2 are served by GLM 5.3 until September 30, 2026; use glm-5.3 going forward.
Do I lose anything with the upgrade? No — GLM 5.3 is the same base model with more post-training, at the same price and with a far larger context window as served here (1M tokens, against 198K for GLM 5.2). The upgrade was applied to every Frontier subscription automatically.
Is there a GLM coding plan? Yes — a Frontier Pool subscription works as a GLM coding plan: unlimited GLM 5.3 through Claude Code, Cline, Roo Code or any OpenAI-compatible coding agent, plus MiniMax M3 on the same subscription.