Zhipu AI / Z.ai

GLM 5.2

Type
language model
Context
1,049K tokens
Max output
131K
Released
13 June 2026
API string
glm-5.2
Weights
open

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
OpenRouter $0.419 $1.316 2 Aug 2026
crof $0.5 $2.2 31 Jul 2026
Weights & Biases $0.76 $2.42 31 Jul 2026
routing-run $0.8 $2.4 31 Jul 2026
LLM Gateway $0.8 $2.55 31 Jul 2026
vultr $0.85 $3.1 31 Jul 2026
lilac $0.9 $3 31 Jul 2026
DeepInfra $0.93 $3 31 Jul 2026
GMI Cloud $0.979 $3.08 31 Jul 2026
aiand $1 $4 31 Jul 2026
ambient z-ai/glm-5.2 $1.05 $4.4 31 Jul 2026
DigitalOcean Gradient AI $1.05 $4.4 31 Jul 2026
Merge Gateway $1.05 $3.3 31 Jul 2026
alibaba-cn $1.1 $3.851 31 Jul 2026
Vercel $1.1 $3.851 31 Jul 2026
AIHubMix $1.127 $3.944 31 Jul 2026
vivgrid $1.2 $4.2 31 Jul 2026
ambient zai-org/GLM-5.2-FP8 $1.2 $4.2 31 Jul 2026
inceptron $1.2 $4.2 31 Jul 2026
wafer.ai $1.2 $4.1 31 Jul 2026
CrossModel $1.2 $4.4 31 Jul 2026
dinference $1.25 $3.89 31 Jul 2026
greenpt $1.254 $5.016 31 Jul 2026
opencode-go $1.4 $4.4 31 Jul 2026
friendli $1.4 $4.4 31 Jul 2026
cloudflare-workers-ai $1.4 $4.4 31 Jul 2026
cline-pass $1.4 $4.4 31 Jul 2026
ofox $1.4 $4.4 31 Jul 2026
synthetic $1.4 $4.4 31 Jul 2026
subconscious $1.4 $4.4 31 Jul 2026
hyper $1.4 $4.4 31 Jul 2026
novita-ai $1.4 $4.4 31 Jul 2026
siliconflow-cn $1.4 $4.4 31 Jul 2026
zai $1.4 $4.4 31 Jul 2026
hpc-ai $1.4 $4.4 31 Jul 2026
Abacus.AI $1.4 $4.4 31 Jul 2026
Baseten $1.4 $4.4 31 Jul 2026
Cloudflare AI Gateway $1.4 $4.4 31 Jul 2026
Cloudflare Workers AI $1.4 $4.4 31 Jul 2026
Databricks $1.4 $4.4 31 Jul 2026
EmpirioLabs AI $1.4 $4.4 31 Jul 2026
Fireworks AI $1.4 $4.4 1 Aug 2026
Hugging Face $1.4 $4.4 31 Jul 2026
Nebius Token Factory $1.4 $4.4 31 Jul 2026
OpenCode Zen $1.4 $4.4 31 Jul 2026
Pioneer $1.4 $4.4 31 Jul 2026
SiliconFlow $1.4 $4.4 31 Jul 2026
Together AI $1.4 $4.4 31 Jul 2026
Umans AI $1.4 $4.4 31 Jul 2026
Z.ai $1.4 $4.4 31 Jul 2026
ZenMux $1.4 $4.5 31 Jul 2026
evroc $1.438 $5.75 31 Jul 2026
Cortecs $1.44 $4.53 31 Jul 2026
neuralwatt $1.45 $4.5 31 Jul 2026
tinfoil $1.5 $5.25 31 Jul 2026
tensorx $1.5 $4.5 2 Aug 2026
berget $1.54 $4.84 31 Jul 2026
unorouter $1.6 $5.029 31 Jul 2026
Scaleway $1.8 $5.5 31 Jul 2026

Only the base tier and only per-token prices are shown. Batch, discounted and cached rates, as well as prices per image or per second of video, do not go into this table: they cannot stand in the same column as a price per million tokens.

A single provider appears several times if it sells this model under different identifiers and at different prices — for example, at different weight precision. The identifier is shown next to the name. Identical offers that arrived from two sources under different spellings are merged into one row.

The date in the last column is the day this price first entered our collection. The price may well be older: before that day we simply were not recording it. It has not changed since — otherwise a new row with a new date would stand in its place.

Measurement results

Task set Result Run conditions Measured by
Arena Score, web development 1,586.29 1,577–1,595 effort: max
Arena Score, programming 1,477.24 1,469–1,486 effort: max
Arena Score in French 1,476.9 1,454–1,500 effort: max
Arena Score in Spanish 1,475.02 1,450–1,500 effort: max
Arena Score, hard prompts 1,472.27 1,466–1,479 effort: max
Arena Score, mathematics 1,471.38 1,453–1,490 effort: max
Arena Score, expert questions 1,469.73 1,456–1,483 effort: max
Arena Score, long queries 1,468.7 1,461–1,476 effort: max
Arena Score in English 1,468.69 1,461–1,476 effort: max
Arena Score, overall 1,464.36 1,459–1,470 effort: max
Arena Score, multi-turn dialogue 1,463.69 1,453–1,474 effort: max
Arena Score in Russian 1,461.6 1,448–1,475 effort: max
Arena Score, creative writing 1,456.24 1,446–1,467 effort: max
Arena Score, instruction following 1,453.73 1,446–1,462 effort: max
GPQA Diamond — graduate-level questions 89.14 % effort: max · with a tuned harness Epoch evaluations
Mock AIME 2024–2025 — olympiad problems 86.38 % effort: max · with a tuned harness Epoch evaluations
SWE-bench Verified — fixing bugs in repositories 78.7 % effort: max · with a tuned harness Epoch evaluations
ARC-AGI — generalising to unseen patterns 77 % · with a tuned harness https://arcprize.org/leaderboard
WeirdML — unusual machine learning tasks 70.12 % effort: max https://htihle.github.io/weirdml.html
FrontierMath, levels 1–3 59.21 % effort: max Epoch evaluations
CursorBench — edits in the editor 54.6 % effort: max https://cursor.com/cursorbench
SimpleQA Verified — factual accuracy 38.1 % effort: max Epoch evaluations
APEX-Agents 35.6 %
FrontierMath, level 4 — research-grade problems 29.27 % effort: max Epoch evaluations
FrontierCode — patches fit to be merged into a project 24.5 % effort: none https://cognition.com/frontiercode
ARC-AGI-2 22.78 %
CritPt — physics problems 20.86 % effort: max
Chess puzzles 16.88 % effort: max Epoch evaluations
Arena Score, agent tasks 0.07 0–0 effort: max