VENICE ATLAS · MODEL DOSSIER · GLM 5.2

Model overview

GLM 5.2

Name: GLM 5.2
Brand: venice.ai
Price: 1.4000 USD
Availability: InStock

GLM-5.2 is the next-generation large language model developed by Zhiyuan AI, featuring significantly enhanced reasoning capabilities, improved instruction following, and support for multiple languages. Supports large context windows for processing extensive text and detailed analysis with fast inference speed.

zai-org-glm-5-2 · venice.ai

Text · Venice catalog$1.40 per 1M tokensPrice on Venice AI

ModalityTextOn Venice AI

Primary price$1.40 per 1M tokensper 1M tokens

Context window1MMax tokens

Benchmarks19Benchmark results

Benchmark results

See the tests behind each score. They use the benchmark’s original scoring scale.

Biology#2033.37 rating Chemistry#2433.37 rating Coding#2933.27 rating Vision#3628.62 rating General intelligence#3344.79 rating Math#841.39 rating Physics#2533.37 rating Reasoning#3644.45 rating Systems#95.72 rating

Agents3 results

PostTrainBenchposttrainbench · agents

34.3%

ReportedOct 2026

Program Benchprogram-bench · agents

63.7%

ReportedOct 2026

Toolathlontoolathlon · agents

48.2%

ReportedOct 2026

Coding5 results

DeepSWEdeepswe · code

46.2%

ReportedOct 2026

DeepSWE 1.1deepswe-1.1 · code

44.0%

ReportedOct 2026

FrontierSWEfrontierswe · code

74.0%

ReportedOct 2026

NL2Reponl2repo · code

48.9%

ReportedOct 2026

SWE-Marathonswe-marathon · code

13.0%

ReportedOct 2026

Math4 results

AIME 2026aime-2026 · math

99.2%

ReportedOct 2026

HMMT 2025hmmt-2025 · math

94.4%

ReportedOct 2026

HMMT Feb 26hmmt-feb-26 · math

92.5%

ReportedOct 2026

IMO-AnswerBenchimo-answerbench · math

91.0%

ReportedOct 2026

Reasoning7 results

CritPTcritpt · reasoning

16.7%

ReportedOct 2026

FrontierCode 1.1frontiercode-1.1 · reasoning

24.5%

ReportedOct 2026

GPQAgpqa · reasoning

91.2%

ReportedOct 2026

Humanity's Last Examhumanity's-last-exam · reasoning

54.7%

ReportedOct 2026

MCP Atlasmcp-atlas · reasoning

76.8%

ReportedOct 2026

SWE-Bench Proswe-bench-pro · reasoning

62.1%

ReportedOct 2026

Terminal-Bench 2.1terminal-bench-2.1 · reasoning

82.7%

ReportedOct 2026

Learn what each benchmark measures ↗

Pricing detail

Input price per 1M tokens: $1.40
Output price per 1M tokens: $4.40
Cache read price per 1M tokens: $0.26
Blended price per 1M tokens (50/50): $2.90

Prices are published by Venice AI. Venice model page ↗

Capabilities

ReasoningFunction callingWeb search

Endpoint metadata

Privacy: private
Quantization: fp8
Maximum completion tokens: 131K
Uncensored: No
Beta: No
Mapping confidence: 100.0%

View model source ↗

All capabilities

Quantization: fp8
Supports E2 E E: No
Supports Vision: No
Supports X Search: No
Optimized For Code: Yes
Supports Log Probs: No
Supports Reasoning: Yes
Supports Web Search: Yes
Supports Audio Input: No
Supports Video Input: No
Default Reasoning Effort: max
Reasoning Effort Options: none, high, max
Supports Multiple Images: No
Supports Response Schema: Yes
Supports Tee Attestation: No
Supports Function Calling: Yes
Supports Reasoning Effort: Yes

Traits

defaultfunction_calling_default

Related models

More from zai-org

z-ai-glm-5-3GLM 5.3$1.75 input per 1M tokens ↗z-ai-glm-5-3-flashGLM 5.3 Flash$0.15 input per 1M tokens ↗z-ai-glm-5v-turboGLM 5V Turbo$1.50 input per 1M tokens ↗