GLM 5.3 on OpenRouter: Availability, Model ID & Setup

Quick answer: As of the August 14, 2026 launch, GLM 5.3 is not yet on OpenRouter — the newest Z.AI model listed there is z-ai/glm-5.2. GLM 5.3 is currently available through the GLM Coding Plan and ZCode, with the public API (and presumably OpenRouter listing) coming soon.

TL;DR

Current Status: OpenRouter Doesn't Have GLM 5.3 Yet

GLM 5.3 launched on August 14, 2026, and as with most new Z.AI releases, the routing ecosystem takes a moment to catch up. Checking OpenRouter's model catalog today shows:

• z-ai/glm-5.2 — available
• z-ai/glm-5.2:free — available
• z-ai/glm-5.2:batch — available
• z-ai/glm-5.3 — not yet listed

Z.AI's own docs confirm the API itself is still rolling out: "The GLM-5.3 API will be available soon." OpenRouter listings normally follow the provider API launch, so expect z-ai/glm-5.3 shortly after — typically within days of the official endpoint going live.

What GLM 5.2 Looks Like on OpenRouter (Pricing Reference)

While you wait, GLM 5.2 remains the GLM entry point on OpenRouter:

That pricing is representative of what GLM 5.3 should cost when listed — Z.AI's official page prices GLM 5.2 at $1.4/$4.4 per 1M, and OpenRouter's rate is in the same band (with typical routing markup).

How to Use GLM 5.3 Right Now (Before OpenRouter)

Option 1: GLM Coding Plan (easiest)

Subscribe at z.ai/subscribe — all Coding Plan users were upgraded to GLM 5.3 on launch day. You can use it inside ZCode, Claude Code, Cline, Kilo Code, OpenCode and 20+ other tools via Z.AI's devpack (npx @z_ai/coding-helper). Off-peak hours cost 50% of standard points.

Option 2: Wait for the API

The glm-5.3 model ID will work on Z.AI's API endpoints once it ships. Z.AI has published the endpoints ahead of launch:

When the API is live, OpenRouter usually adds the model within days — the routing community tracks provider releases closely.

Option 3: GLM 5.2 on OpenRouter now

If you need an OpenRouter-compatible GLM today, z-ai/glm-5.2 works with any OpenAI-compatible client. Note the API difference when you later switch to 5.3: GLM 5.3 always has thinking enabled (reasoning_effort: low/high/max) — thinking.type: "disabled" will fail.

What GLM 5.3 Changes vs GLM 5.2 (For Your OpenRouter Migration)

When z-ai/glm-5.3 lands, your request body changes in one important way:

{
  "model": "z-ai/glm-5.3",
  "thinking": { "type": "enabled" },
  "reasoning_effort": "max"
}

• reasoning_effort: low | high | max (default max; max recommended for coding).
• If your app sends thinking.type: "disabled" (valid for 5.2), remove it — GLM 5.3 rejects it.
• Everything else — streaming, function calling, 1M context — carries over.

FAQ

Is GLM 5.3 available on OpenRouter? Not as of August 14, 2026. The newest GLM on OpenRouter is z-ai/glm-5.2. Expect z-ai/glm-5.3 after Z.AI's public API goes live.

What is the OpenRouter model ID for GLM 5.3? Not listed yet. By convention it should be z-ai/glm-5.3, matching the pattern of z-ai/glm-5.2.

How much does GLM 5.2 cost on OpenRouter? About $1.19 per 1M input and $3.74 per 1M output; a free variant (z-ai/glm-5.2:free) also exists.

Can I use GLM 5.3 with my OpenRouter-based tools? Not through OpenRouter yet. Use the GLM Coding Plan with ZCode/Claude Code/Cline now, or wait for the API + routing listing.

Will GLM 5.3 work with thinking disabled on OpenRouter? No — GLM 5.3 requires thinking enabled everywhere, including any future OpenRouter listing.

Sources

• Z.AI: GLM-5.3: Frontier Coding with Emergent Cyber Capabilities (August 14, 2026)
• Z.AI Docs: GLM-5.3
• OpenRouter Models API

Last updated: August 14, 2026