How much does GLM 5.3 cost?

GLM 5.3 costs $0.03 per million input tokens and $4.80 per million output tokens. For a typical conversation (~2,000 tokens), that's approximately $0.005 per message.

Is GLM 5.3 open source?

No, GLM 5.3 is a proprietary model by z-ai.

How does GLM 5.3 compare to Qwen2.5 Coder 14B Instruct?

GLM 5.3 has an average score of 64.1 while Qwen2.5 Coder 14B Instruct scores 64.1. Qwen2.5 Coder 14B Instruct slightly outperforms GLM 5.3 overall. See full comparison →

Home/Models/GLM 5.3

GLM 5.3

Name: GLM 5.3 · benchmarks, pricing and providers
Creator: BenchGecko

by z-ai · Released Aug 2026

1M Context

64.1

avg score

Rank #66

Compare

Better than 79% of all models

Context

1.0M tokens (~524 books)

Input $/1M

$0.03

Output $/1M

$4.80

Type

text

License

Proprietary

Benchmarks

22 tested

Data as of 2026-10-15

About

GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks. It supports text input and output with a 1M-token context window, and improves...

Tested on 22 benchmarks · BenchGecko score 64.1. Top scores: Artificial Analysis · GPQA Diamond (91.7%), OTIS Mock AIME 2024-2025 (91.1%), GPQA diamond (87.9%).

Looking for similar performance at lower cost?
GLM 5.2 scores 64.3 (100% as good) at $0.02/1M input · 33% cheaper

Capabilities

coding

53.7

#78 globally

math

63.0

#78 globally

knowledge

54.2

#100 globally

agentic

56.6

#17 globally

general

55.0

#24 globally

speed

56.3

#16 globally

Benchmark Scores

Compare All

Tested on 22 benchmarks · Ranked across 6 categories

Score Distribution (all 319 models)

0255075100

▲ You are here

codingCompare coding →

WeirdML

Unusual and adversarial machine learning challenges. Tests robustness of reasoning about edge cases in ML systems.

75.4·

Deepswe

69.0·

Frontiercode

40.1·

mathCompare math →

OTIS Mock AIME 2024-2025

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

91.1·

FrontierMath-Tiers-1-3-v2-Private

68.8·

FrontierMath-Tier-4-v2-Private

29.3·

knowledgeCompare knowledge →

GPQA diamond

Graduate-level science questions written by PhD experts. Diamond subset contains questions where experts disagree, testing deep understanding.

87.9·

SimpleQA Verified

Simple factual questions with verified correct answers. Tests accuracy of basic knowledge retrieval. Low scores indicate hallucination.

41.0·

Chess Puzzles

Tactical chess puzzles testing pattern recognition and multi-move calculation. Measures strategic reasoning ability.

33.7·

Quick compare:

vs Qwen2.5 Coder 14B Instruct

vs Gemini 3.1 Pro Preview

vs Meta Llama 3 8B

Excellent (85+) Good (70-85) Average (50-70) Below (<50)

Model Family · z-ai GLM 5.3

GLM 5.3Aug 2026

64.1

$0.03/M in1.0Mctx22 benchmarks

GLM 5.3 (batch)Aug 2026

$0.45/M in(+0.42)1.0Mctx

GLM 5.3 FlashAug 2026

50.5-13.6

$0.15/M in(-0.30)1.0Mctx23 benchmarks

GLM 5.3 Flash (batch)Aug 2026

$0.06/M in(-0.09)1.0Mctx

GLM 5.3 FlashXSep 2026

$0.37/M in(+0.31)1.0Mctx

GLM 5.3 PrimeSep 2026

$2.80/M in(+2.43)1.0Mctx(-49K)

See the full GLM 5.3 family →

Recently Happened

GLM 5.3 added

Aug 19, 2026

Similar Models

Qwen2.5 Coder 14B Instruct

Alibaba

64.1n/a

Gemini 3.1 Pro Preview

Google DeepMind

64.1$2.00/1M

Meta Llama 3 8B

Frequently Asked Questions

GLM 5.3 is a proprietary text AI model by z-ai, released in August 2026. It has an average benchmark score of 64.1. Context window: 1M tokens.

Benchmarks

Artificial Analysis · GPQA Diamond OTIS Mock AIME 2024-2025 GPQA diamond Artificial Analysis · Long Context Reasoning Dtbench

z-ai · Provider z-ai · Economy All Models Compare Models Pricing Developers · API Context Window · Glossary

Key facts · as of 2026-10-15

GLM 5.3 by z-ai. BenchGecko score 64.1, rank 66 of 319 scored models (normalized average of public benchmark scores).
List price $0.0300 input · $4.80 output per 1M tokens (as of 2026-10-15).
Sold by 43 providers (as of 2026-10-15): Relace $0.0630 in / $4.40 out · Wafer $0.0790 in / $4.80 out · InferenceNet (fp4) $0.0800 in / $4.80 out · Morph (fp8) $0.0800 in / $6.00 out · Makora (fp4) $0.14 in / $4.40 out · and 38 more. Every provider
Gecko Tests: Who Are You B (Knows who made it) · Tokenizer Tax E (90% more tokens outside English).

How to cite · data as of 2026-10-15

GLM 5.3 · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-15. https://benchgecko.ai/model/glm-5-3

Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP

GLM 5.3

Frequently Asked Questions

Related Models

Benchmarks

Related Pages

Key facts · as of 2026-10-15

How to cite · data as of 2026-10-15