All benchmarks
Zhipu AI · China · Released 2025-07

GLM-4.5

Zhipu's unified reasoning + coding + agent model, positioned as a Claude-class open alternative.

Access
open-weights
Context
128,000 tokens
Modalities
text, vision
Languages
40+

Benchmark scores

MMLU
84.6%
Massive Multitask Language Understanding
GPQA
79.1%
Graduate-level science QA (Diamond)
HumanEval
90.6%
Python code completion
SWE-bench
64.2%
Real GitHub issue resolution (Verified)
LiveCodeBench
N/A
Contamination-free coding
AIME
N/A
American Invitational Mathematics Exam
MMMU
N/A
Multimodal college-level reasoning
MATH
N/A
Competition-level math word problems

Scores as reported by the vendor or leading public leaderboards. "N/A" means the score has not been publicly disclosed for this metric.

Explain like I'm 5

An open Chinese model that's good at reasoning, coding and being an agent.

Key features
  • Reasoning mode
  • Tool use
  • Open weights
Strengths
  • Balanced across benchmarks
  • Truly open license
Limitations
  • Newer ecosystem

Best for

Open agentsEnterprise coding

Related models