GLM 5.3 Is Officially Announced

GLM 5.3

GLM 5.3 is the new Zhipu AI (Z.ai) flagship large language model: 1M-token context, 128K maximum output, and top open-source coding benchmarks.

Start a GLM 5.3 Chat GLM 5.3 Benchmarks GLM 5.3 Specs

1 Credit

What Is the GLM 5.3 Model?

GLM 5.3 is the new flagship large language model from Zhipu AI (Z.ai), announced in August 2026, built on the previous flagship base with large-scale long-horizon post-training.

The new Z.ai flagship

Z.ai newest flagship base model: 1M-token context window, 128K maximum output, and text-in, text-out modality.

Same base, far more training

It uses the previous flagship base and adds tens of times more long-horizon post-training across real compute clusters, storage, docs, and codebases; Z.ai says coding feel improves by about 50%.

An honest release tracker

The GLM 5.3 model API is coming soon, so this page separates what Z.ai has confirmed (specs and benchmarks) from what is still pending (API ID, pricing, weights, and license).

GLM 5.3 Release Status: Confirmed Facts and Pending Details

Z.ai has confirmed GLM 5.3 specs and benchmarks. API access, pricing, and open weights are still pending announcements.

Confirmed by Z.ai

Verified facts

Context window: 1M tokens
Maximum output: 128K tokens
Modality: Text in / text out
GLM Coding Plan: Live
Model API: Coming soon

Pending official details

Tracked expectations

API pricing: Unannounced
Open weights: Expected, unconfirmed
Parameters: Unpublished
Official model ID: Expected glm-5.3

Status is checked against Z.ai public communications as of August 2026; details not marked as confirmed remain unverified until Z.ai announces them.

Benchmarks / 02

GLM 5.3 benchmarks vs GLM 5.2: use them as a starting point

Z.ai reports large GLM 5.3 benchmark gains, but a score only applies to the published harness and date.

Z.ai-reported result

Terminal-Bench 3.0

5.328.3

5.24.6

Z.ai reports 28.3 vs 4.6 for the previous flagship on Terminal-Bench 3.0 — the largest reported jump.

Z.ai-reported result

DeepSWE v1.1

5.366.9

5.246.2

Z.ai reports 66.9 vs 46.2 for the previous flagship on DeepSWE v1.1, a long-horizon engineering benchmark.

Z.ai-reported result

Agents' Last Exam

5.328.5

5.223.8

Z.ai reports 28.5 vs 23.8 for the previous flagship on Agents' Last Exam, covering agentic knowledge and reasoning.

Z.ai-reported result

CyberGym

5.384.5

GPT-5.6 Sol83.6

Z.ai reports 84.5 on CyberGym, slightly above GPT-5.6 Sol at 83.6 — a published result, not a guarantee for your security workflow.

Capabilities / 03

GLM 5.3 capabilities

GLM 5.3 supports the standard developer stack for agentic and long-horizon work, with 1M-token context for project-scale tasks.

Thinking modes

Multiple thinking modes balance reasoning depth against latency, matching the task at hand.

Function calling

Strong tool-calling support for external integrations, with your application validating and executing every requested action.

Streaming output

Real-time streaming responses for interactive chat and agent loops.

Context caching

Smart caching optimizes long conversations and repeated context, cutting cost on long-horizon sessions.

Structured output

JSON and other structured output formats make system integration predictable.

MCP support

The model can call external MCP tools and data sources, extending agent workflows beyond chat.

Release timeline / 04

GLM 5.3 release timeline and news

The milestones that led to GLM 5.3, and what still needs to land, updated as Z.ai publishes official news.

Jun 16, 2026 — GLM 5.2 launches

GLM 5.2 launched for long-horizon tasks with a 1M-token context window and effort-level control.

Aug 2026 — GLM 5.3 announced

Z.ai published the official model page and benchmark results, naming it the new flagship built on the GLM 5.2 base.

Aug 2026 — Coding Plan goes live

The Z.ai Coding Plan fully rolled out the model for coding tools like Claude Code, Cline, and OpenCode.

Next — GLM 5.3 model API

Z.ai says the API is coming soon. The official model ID, limits, and pricing will be updated here.

Next — open weights and license

GLM 5.3 download and license details are expected to follow the MIT precedent of its predecessor, pending official confirmation.

Where GLM 5.3 news breaks first

Follow the Z.ai blog, the zai-org GitHub repository, and the bigmodel docs for the first API announcements.

Who it is for / 05

Who uses GLM 5.3?

GLM 5.3 is aimed at teams doing long-horizon engineering, agentic workflows, and security review.

Application developers

Developers building coding assistants and IDE workflows benefit from the reported 50% coding feel gain and the 1M-token context for whole-repository work.

Agent builders

Agent teams use thinking modes, function calling, streaming, and MCP to build long-running workflows with clear stop conditions.

Security researchers

It shows strength in white-box code review and vulnerability discovery (CyberGym 84.5), useful as an assistant layer experts still validate.

Long-horizon engineering teams

Teams working across tens of thousands of lines and dozens of files use it to plan refactors, fix bugs, and deliver projects with minimal hand-holding.

API and platform teams

Platform teams track API pricing and limits to budget for context caching and structured output before the official API goes live.

Open-source community

Self-hosters watch for open weights to plan hardware, quantization, and license review, following the previous MIT precedent.

How to use GLM 5.3 / 06

How to use GLM 5.3

GLM 5.3 is available in glm5.app Chat today, with the official model API expected to follow.

Start a GLM 5.3 chat

Open Chat on glm5.app, describe the problem, constraints, and desired outcome, then refine the response with your code.

Subscribe to the GLM Coding Plan

For coding-tool integrations today, the Z.ai Coding Plan already serves it in tools like Claude Code, Cline, and OpenCode.

Integrate when the API goes live

For glm5.app API access, use model ID `glm-5.3` with the OpenAI-compatible base URL `https://glm5.app/api/v1`.

Integration specimen

Start with the GLM 5.3 API

The official GLM 5.3 model API is coming soon from Z.ai. glm5.app exposes an OpenAI-compatible Chat Completions endpoint today, and the official model ID will be added here when published.

Open API documentation

Request previewcurl

curl https://glm5.app/api/v1/chat/completions \
  -H "Authorization: Bearer $GLM5_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "glm-5.3",
    "messages": [{
      "role": "user",
      "content": "Plan this refactor and list the rollback steps."
    }]
  }'

GLM 5.3 FAQ

Clear GLM 5.3 answers about release date, pricing, benchmarks, Hugging Face, parameters, weights, the coding plan, and the upgrade from 5.2.

What is GLM 5.3?

The new flagship large language model from Z.ai, announced in August 2026.

When is the GLM 5.3 release date?

The official model page and benchmarks arrived in August 2026. The GLM Coding Plan already serves it, and the model API is expected to go live shortly.

What is GLM 5.3 pricing?

API pricing has not been published. For reference, the previous flagship costs about $1.40 input and $4.40 output per million tokens.

Is GLM 5.3 on Hugging Face?

Weights are not on Hugging Face yet. Z.ai open-sourced the previous model under the MIT license, so a release there is widely expected, but it has not been published as of August 2026.

How many parameters does GLM 5.3 have?

Parameter counts are unpublished; the official page lists 1M-token context and 128K maximum output. Totals likely follow the GLM 5 family MoE architecture.

Will GLM 5.3 weights be open source?

Weights are expected under the MIT license, following the previous release on Hugging Face. Z.ai has not confirmed the license yet, so treat open-weights details as unconfirmed.

What is the GLM 5.3 coding plan?

The Z.ai Coding Plan — a subscription for AI coding tools like Claude Code, Cline, and OpenCode — has fully rolled out the new model.

How can I use GLM 5.3 now?

Chat with it free in glm5.app Chat today; the GLM Coding Plan also serves it. The public model API is expected to go live soon, and this page will update with the official ID and limits when announced.

Is glm5.3 the same as GLM 5.3?

Yes — glm5.3, GLM-5.3, and GLM 5.3 are spelling variants of the same model.

Start here

Start with GLM 5.3 today

Chat with GLM 5.3 free on glm5.app, and track the official API, pricing, and open-source news here.

Start a GLM 5.3 Chat Explore GLM 5.2 View Pricing

GLM 5.3

Create with GLM 5

What Is the GLM 5.3 Model?

The new Z.ai flagship

Same base, far more training

An honest release tracker

GLM 5.3 Release Status: Confirmed Facts and Pending Details

Verified facts

Tracked expectations

GLM 5.3 benchmarks vs GLM 5.2: use them as a starting point

Terminal-Bench 3.0

DeepSWE v1.1

Agents' Last Exam

CyberGym

GLM 5.3 capabilities

Thinking modes

Function calling

Streaming output

Context caching

Structured output

MCP support

GLM 5.3 release timeline and news

Jun 16, 2026 — GLM 5.2 launches

Aug 2026 — GLM 5.3 announced

Aug 2026 — Coding Plan goes live

Next — GLM 5.3 model API

Next — open weights and license

Where GLM 5.3 news breaks first

Who uses GLM 5.3?

Application developers

Agent builders

Security researchers

Long-horizon engineering teams

API and platform teams

Open-source community

How to use GLM 5.3

Start a GLM 5.3 chat

Subscribe to the GLM Coding Plan

Integrate when the API goes live

Start with the GLM 5.3 API

GLM 5.3 FAQ

What is GLM 5.3?

When is the GLM 5.3 release date?

What is GLM 5.3 pricing?

Is GLM 5.3 on Hugging Face?

How many parameters does GLM 5.3 have?

Will GLM 5.3 weights be open source?

What is the GLM 5.3 coding plan?

How can I use GLM 5.3 now?

Is glm5.3 the same as GLM 5.3?

Start with GLM 5.3 today