Model

GLM-4-Plus

CLOSED

Zhipu AI · GLM-4 family · released Aug 29, 2024

Closed flagship unveiled at KDD 2024; official docs claim AlignBench 83.2 (99% of GPT-4o) and MMLU 86.8 (98% of GPT-4o), i.e. near-parity language/instruction-following/long-text performance.

ReasoningCodingVisionFunction callingTool useAgentic

1115.1

Elo · rank #189

Parameters

Undisclosed

Active params

—

Context

128K tokens

Architecture

Undisclosed architecture (assumed dense transformer)

License

Proprietary

Languages

—

API price (in/out)

$0.69 / $0.69

Modalities

text

Benchmark results

Bar shows position within the tracked field; marker = field best

GPQA DiamondReasoning50.7%#186

best: GPT-5.6 · 94.6%

IFEvalReasoning79.5%#91

best: Gemma 4 26B A4B · 98.5%

LiveCodeBenchCoding45.8%#123

best: DeepSeek-V4-Pro (Think Max) · 93.5%

MATH-500Math74.2%#80

best: GPT-5 · 99.4%

MMLUKnowledge86.8%#39

best: OpenAI o3 · 92.9%

Run it locally

Closed weights — available via API only. No local deployment.

Input / M tok

$0.69

Output / M tok

$0.69

GLM-4 family

Elo progression across releases

GLM-4-9B-ChatJun 2024629.9 GLM-4V-9BJun 2024675.7 GLM-4-PlusAug 20241115.1

Resources

DOCSdocs.bigmodel.cn APIAvailable via Zhipu AI API

API price $0.69/$0.69 · each benchmark row carries its own source badge (see methodology)