Home/Models/GLM 5
z-ai logo

GLM 5

by z-ai · Released Feb 2026

Open Source
69.5
avg score
Rank #39
Compare
Better than 83% of all models
Context
203K tokens (~101 books)
Input $/1M
$0.60
Output $/1M
$1.92
Type
text
License
Open Source
Benchmarks
28 tested
Data updated today
About

GLM-5 is Z.ai’s flagship open-source foundation model engineered for complex systems design and long-horizon agent workflows. Built for expert developers, it delivers production-grade performance on large-scale programming tasks, rivaling leading...

Tested on 28 benchmarks with 57.6% average. Top scores: Chatbot Arena Elo — Overall (1455.6%), Chatbot Arena Elo — Coding (1441.0%), OpenCompass — AIME2025 (95.8%).

Looking for similar performance at lower cost?
Gemma 4 31B scores 68.2 (98% as good) at $0.13/1M input · 78% cheaper
Capabilities
coding
64.6
#30 globally
reasoning
46.1
#54 globally
math
55.6
#61 globally
knowledge
53.6
#82 globally
language
75.4
#58 globally
Benchmark Scores
Compare All
Tested on 28 benchmarks · Ranked across 6 categories
Score Distribution (all 233 models)
0255075100
▲ You are here
OpenCompass — LiveCodeBenchV6

OpenCompass Live Code Bench v6. Fresh competitive programming problems to evaluate code generation without memorization.

86.2
LiveBench — Coding

Regularly refreshed coding problems that avoid data contamination. New problems added monthly to prevent memorization.

73.6
SWE-Bench verified

Real-world software engineering tasks from GitHub issues. Models must diagnose bugs and write patches that pass test suites. Human-verified subset of SWE-bench.

72.1
LiveBench — Reasoning

Regularly refreshed reasoning problems testing logical deduction, spatial reasoning, and analytical thinking.

69.1
LiveBench — Data Analysis

Fresh data analysis tasks testing ability to interpret tables, charts, and statistical data.

67.9
ARC-AGI

Abstraction and Reasoning Corpus. Tests fluid intelligence through novel visual pattern recognition puzzles. Core measure of general intelligence.

44.7
OpenCompass — AIME2025

OpenCompass evaluation on AIME 2025 problems. Tests mathematical reasoning on fresh competition problems.

95.8
LiveBench — Mathematics

Regularly updated math problems that test numerical reasoning, algebra, calculus, and combinatorics.

83.5
OTIS Mock AIME 2024-2025

Mock AIME (American Invitational Mathematics Exam) problems from OTIS. Tests mathematical competition performance.

80.0
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Links
Documentation
Community
BenchGecko API
glm-5
Specifications
  • Typetext
  • Context203K tokens (~101 books)
  • ReleasedFeb 2026
  • LicenseOpen Source
  • StatusActive
  • Cost / Message~$0.003
Available On
z-ai logoz-ai$0.60
Share & Export
Tweet
GLM 5 is an open-source text AI model by z-ai, released in February 2026. It has an average benchmark score of 69.5. Context window: 203K tokens.