Skip to content

Z.ai mit Released Sep 30, 2025 Updated

GLM-4.6

ToolsGuide Scores, rankings, test results and price.

Model summary

  • Text #124 / 416

    339 / 1,000

    ToolsGuide Score· votes 345, tests 333

  • Coding #119 / 147

    239 / 1,000

    ToolsGuide Score· votes 244, tests 232

  • Price per 1M tokens

    In $0.50· Out $2.00

    $0.50

    Input · median text model $0.43

Summary

GLM-4.6 is Z.ai's model, released Sep 30, 2025. It's ranked in 2 categories and does best at text: #124 of 416, with a ToolsGuide Score of 339 against a median of 235.

It also ranks #119 of 147 for coding (239).

It costs $0.50 per million input tokens and $2.00 per million output tokens, close to the median text model ($0.43).

It has published results on 7 tests.

Specifications

Creator
Z.ai
Licence
mit
Released
Sep 30, 2025
Context window
203K tokens
Input price
$0.50 / 1M
Output price
$2.00 / 1M
Tests reported
7
Settings ranked
glm-4.6, GLM-4.6

Model comparison

Against the ten best text models.

Text score

ToolsGuide Score out of 1,000 · Higher is better

Coding score

ToolsGuide Score out of 1,000 · Higher is better

Input price

USD per 1M input tokens · Lower is better

Leaderboards

Where GLM-4.6 sits among the top 20 in each category.

Text

#124 of 416 · ToolsGuide Score out of 1,000, where 1,000 is a flawless model

Full text leaderboard →

Coding

#119 of 147 · ToolsGuide Score out of 1,000, where 1,000 is a flawless model

Full coding leaderboard →

By task

Scored from votes on each kind of prompt alone, on the same 1,000-point scale.

Score against price

Text · the 10 priced models ranked closest · up and to the left is better value

Against its nearest rivals

Text prompts by kind · ToolsGuide Score from votes · hover for values

Text

glm-4.6 · hover a bar for its rank

Coding

GLM-4.6 · hover a bar for its rank

Test results

Every test GLM-4.6 has a published result on, out of 100%, with its rank among the models that took it.

  • GPQA

    #96 / 256

    Graduate-level science questions written to be hard to look up.

    81.0%

    Self-reported

  • AIME 2025

    #29 / 123

    Competition maths problems from the 2025 American Invitational Mathematics Exam.

    93.9%

    Self-reported

  • SWE-Bench Verified

    #78 / 120

    Fixing real bugs in real GitHub projects, checked by the projects' own tests.

    68.0%

    Self-reported

  • Humanity's Last Exam

    #78 / 108

    Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions

    17.2%

    Self-reported · In the text score· In the vision score

  • BrowseComp

    #56 / 68

    Finding hard-to-locate facts by browsing the web.

    45.1%

    Self-reported · In the agents score

  • LiveCodeBench v6

    #20 / 67

    LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.

    82.8%

    Self-reported

  • Terminal-Bench

    #8 / 25

    Terminal-Bench is a benchmark for testing AI agents in real terminal environments. It evaluates how well agents can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities. The benchmark consists of a dataset of ~100 hand-crafted, human-verified tasks and an execution harness that connects language models to a terminal sandbox.

    40.5%

    Self-reported