GLM-5.3

Leaderboards

Where GLM-5.3 sits among the top 20 in each category.

By task

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

Test results

Every test GLM-5.3 has a published result on, out of 100%, with its rank among the models that took it.

• Humanity's Last Exam

  #5 / 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

  62.5%

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

• DeepSWE 1.1

  #27 / 50

  DeepSWE 1.1 evaluates software engineering agents using the mini-swe-agent harness.

  66.9%

  Self-reported · In the coding score

• Terminal-Bench 2.1

  #7 / 46

  Terminal-Bench 2.1 is an updated release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal. It evaluates how well models handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, data science workflows, and security tasks.

  88.2%

  Self-reported

• Toolathlon

  #8 / 42

  Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.

  73.0%

  Self-reported · In the agents score

• Terminal-Bench 4.0

  #12 / 28

  Terminal-Bench 4.0 evaluates AI agents on real-world work in terminal and command-line containerized environments across 66 tasks, with emphasis on science-adjacent and frontier engineering problems. Relative to earlier Terminal-Bench releases, 4.0 increases timeouts and adaptively raises RAM/CPU on selected tasks to reduce harness and resource confounds.

  41.8%

  Self-reported · In the coding score· In the agents score

• NL2Repo

  #6 / 25

  NL2Repo evaluates long-horizon coding capabilities including repository-level understanding, where models must generate or modify code across entire repositories from natural language specifications.

  58.0%

  Self-reported · In the coding score

• Agents' Last Exam

  #17 / 24

  Agents' Last Exam is a challenging benchmark for AI agents on hard, long-horizon tasks that test sustained reasoning, planning, and tool use, reported with and without tool access.

  28.5%

  Self-reported · In the agents score

• AutomationBench

  #7 / 24

  AutomationBench is a tool-use benchmark that evaluates AI agents on automating real-world workflows, testing their ability to orchestrate tools and complete multi-step automation tasks.

  48.2%

  Self-reported · In the agents score

• CyberGym

  #5 / 21

  CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.

  84.5%

  Self-reported

• Program Bench

  #17 / 18

  Program Bench evaluates code-generation agents by asking them to recreate a program's behavior from only a compiled binary and documentation. It spans 200 tasks from small CLI tools to large systems such as FFmpeg and SQLite, with submissions judged against more than 248,000 fuzz-generated behavioral tests.

  19.0%

  Self-reported · In the coding score

• FrontierSWE

  #3 / 16

  FrontierSWE measures whether an agent can complete open-ended technical projects at the scale of hours to tens of hours, spanning systems optimization, large-scale code construction, and applied ML research. Performance is reported as a dominance score, where higher is better.

  78.1%

  Self-reported

• GDPval-AA

  #3 / 14

  GDPval-AA evaluates AI agents on economically valuable professional knowledge-work tasks and reports performance as an Elo score.

  59.0%

  Self-reported

• SWE-Marathon

  #6 / 10

  SWE-Marathon is an ultra-long-horizon software engineering benchmark covering tasks such as building compilers, optimizing kernels, and developing production-grade services. It measures whether agents can sustain quality across extremely long engineering trajectories.

  42.5%

  Self-reported

• PostTrainBench

  #2 / 9

  PostTrainBench evaluates a model's ability to autonomously post-train base models. Given pretrain-only base models, the agent must complete the full pipeline of data synthesis, training, evaluation, and iteration within a time budget, scored across downstream benchmarks such as AIME2025, BFCL, GPQA Main, GSM8K, and HumanEval.

  39.8%

  Self-reported

• ExploitBench

  #4 / 8

  ExploitBench is a cybersecurity benchmark that evaluates a model's ability to discover and exploit software vulnerabilities, reported as the fraction of challenges where the model captures the target (Cap%).

  54.4%

  Self-reported

• ExploitGym

  #6 / 8

  ExploitGym is a large-scale, realistic benchmark built from real-world vulnerabilities across userspace programs, Google's V8 engine, and the Linux kernel. Given a vulnerability and a proof-of-vulnerability input, agents must craft a working end-to-end exploit that achieves unauthorized code execution.

  15.0%

  Self-reported

• Terminal-Bench 3.0

  #2 / 4

  Terminal-Bench 3.0 is a release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal on real-world, end-to-end tasks.

  28.3%

  Self-reported