Ctrl K
Back to models

Z.AI

GLM-5.3-Flash

  • Released
  • Open weights
  • MIT
Compare
Intelligence Index

Max setting

41.8#18 of 83measured models
Measured · Artificial Analysis v4.3.2
API price from
$0.15 in / $0.50 out / 1M tokZ.AI · direct API
Context window
1.05M tok

Performance profile

ReasoningRank #17 of 107
91.2%
GPQA Diamond
CodingRank #28 of 94
51.6%
SciCode
Cost efficiencyRank #20 of 67
147.4pts / $
Intelligence ÷ eval cost

Capabilities

Context window

Modalities

Reasoning modes

Avg reasoning tokens / task — Artificial Analysis Intelligence Index v4.3.2: weighted average reasoning tokens per benchmark task, not a token limit. Source (opens in a new tab)

Integration

Reasoning trade-offs

Source (opens in a new tab)
Intelligence Index v4.3.2
GLM-5.3-Flash
Cost per task (USD)
GLM-5.3-FlashMax
Intelligence Index
41.8
Cost / task
$0.2836
Evaluation time
12.1m

Pricing & access

API route
Z.AIAPI
Tier & contextStandard

USD per 1M tokens

Input
$0.15
Output
$0.50
Cached input
$0.03

Run locally

GGUFQ8_0
Context
Estimated memory

~349.6 GiB

RAM + VRAM

Estimate assumptions

One text chat at the selected context, with a 16-bit cache and app allowance. Operating system and image processing are extra; optimized caches can use less.

384 GiB RAMwithout GPU offload

Memory budgetOne text chat
Weights
317.6 GiB
Context + state
0.2 GiB16-bit cache
App allowance
31.8 GiB
Download
341 GB8 files · keep together+1.1 GB for images (opens in a new tab)
Disk space
~382 GBIncludes image file
Available software
Unsloth Desktop (opens in a new tab)llama.cpp (opens in a new tab)
Use Unsloth Desktop or the linked llama.cpp build (opens in a new tab)
Get GGUF (opens in a new tab)
Images need extra memoryvLLM setup (opens in a new tab)

Benchmarks

Full leaderboardSource (opens in a new tab)

Humanity’s Last Exam

Frontier reasoning

GLM-5.3-Flash · Max#25 of 126

  1. 1Claude Opus 5.5Max61.4%
  2. 2Claude Fable 5.1Max59.1%
  3. 3Gemini 4 ArgonHigh57.1%
  5. 22Qwen3.8 2.4T A95BReasoning42.4%
  6. 23GLM-5.3Max42.3%
  7. 24DeepSeek-V4-Pro-0813Max41.0%
  8. 25GLM-5.3-FlashMax39.9%
  9. 26Ling-3.1-flashReasoning39.4%
  10. 27DeepSeek-V4.1-FlashMax39.2%
  11. 28MiniMax-M3Reasoning39.0%

Peers use their best measured result on this benchmark.

Alternatives

Each model at its highest-scoring measured setting · Intelligence Index v4.3.2 and its evaluation cost per task.