Shared benchmarks

10
Shared benchmarks for GLM-4.7 and GLM-5
BenchmarkGLM-4.7GLM-5
  • 52.0%
  • 62.0%
Evaluation details for BrowseComp

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

GLM-4.7

Record for GLM-4.7 on BrowseComp BrowseComp

GLM-5

Record for GLM-5 on BrowseComp BrowseComp
  • 85.7%
  • 86.0%
Evaluation details for GPQA Diamond

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

GLM-4.7

Record for GLM-4.7 on Graduate-Level Google-Proof Q&A Diamond

GLM-5

Record for GLM-5 on Graduate-Level Google-Proof Q&A Diamond
  • 83.3%
  • 87.8%
Evaluation details for GPQA Diamond — Epoch AI harness

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

GLM-4.7

Record for GLM-4.7 on Graduate-Level Google-Proof Q&A Diamond — Epoch AI harness

GLM-5

Record for GLM-5 on Graduate-Level Google-Proof Q&A Diamond — Epoch AI harness
  • 80.1%
  • 83.3%
Evaluation details for GPQA Diamond — Vals AI harness

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

GLM-4.7

Record for GLM-4.7 on Graduate-Level Google-Proof Q&A Diamond — Vals AI harness

GLM-5

Record for GLM-5 on Graduate-Level Google-Proof Q&A Diamond — Vals AI harness
  • 82.7%
  • 86.0%
Evaluation details for MMLU-Pro Original — Vals AI harness

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

GLM-4.7

Record for GLM-4.7 on MMLU-Pro Original — Vals AI harness

GLM-5

Record for GLM-5 on MMLU-Pro Original — Vals AI harness
  • 66.7%
  • 73.3%
Evaluation details for SWE-bench Multilingual

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

GLM-4.7

Record for GLM-4.7 on SWE-bench Multilingual

GLM-5

Record for GLM-5 on SWE-bench Multilingual
  • 73.8%
  • 77.8%
Evaluation details for SWE-bench Verified

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

GLM-4.7

Record for GLM-4.7 on SWE-bench Verified

GLM-5

Record for GLM-5 on SWE-bench Verified
  • 69.4%
  • 71.4%
Evaluation details for SWE-bench Verified — Vals AI mini-swe-agent

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

GLM-4.7

Record for GLM-4.7 on SWE-bench Verified — Vals AI mini-swe-agent

GLM-5

Record for GLM-5 on SWE-bench Verified — Vals AI mini-swe-agent
  • 38.2%
  • 49.4%
Evaluation details for Terminal-Bench 2.0 — Vals AI Terminus 2

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

GLM-4.7

Record for GLM-4.7 on Terminal-Bench 2.0 — Vals AI Terminus 2

GLM-5

Record for GLM-5 on Terminal-Bench 2.0 — Vals AI Terminus 2
  • 23.8%
  • 39.2%
Evaluation details for Toolathlon Original

The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry.

GLM-4.7

Record for GLM-4.7 on Toolathlon Original

GLM-5

Record for GLM-5 on Toolathlon Original