Zhipu AI / Z.ai

GLM 5

Type
language model
Context
205K tokens
Max output
131K
Released
11 February 2026
Knowledge cutoff
January 2025
API string
glm-5
Weights
open

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
NanoGPT zai-org/glm-5 $0.3 $2.55 31 Jul 2026
crof $0.48 $1.9 31 Jul 2026
ZenMux $0.58 $2.6 31 Jul 2026
302.AI $0.6 $2.6 31 Jul 2026
CrossModel $0.6 $3 31 Jul 2026
DeepInfra $0.6 $2.08 31 Jul 2026
GMI Cloud $0.6 $1.92 31 Jul 2026
Kilo Code $0.72 $2.3 31 Jul 2026
LLM Gateway $0.72 $2.3 31 Jul 2026
dinference $0.75 $2.4 31 Jul 2026
meganova $0.8 $2.56 31 Jul 2026
OpenRouter openrouter/z-ai/glm-5 $0.8 $2.56 31 Jul 2026
hyper $0.86 $2.784 1 Aug 2026
alibaba-cn $0.86 $3.15 31 Jul 2026
Baseten $0.95 $3.15 31 Jul 2026
FastRouter $0.95 $3.15 31 Jul 2026
OpenRouter z-ai/glm-5 $0.95 $2.55 31 Jul 2026
SiliconFlow $0.95 $2.55 31 Jul 2026
opencode-go $1 $3.2 31 Jul 2026
novita-ai $1 $3.2 31 Jul 2026
siliconflow-cn $1 $3.2 31 Jul 2026
zai $1 $3.2 31 Jul 2026
tensorx $1 $3.2 2 Aug 2026
Abacus.AI $1 $3.2 31 Jul 2026
DigitalOcean Gradient AI $1 $3.2 31 Jul 2026
Google Vertex AI $1 $3.2 31 Jul 2026
Hugging Face $1 $3.2 31 Jul 2026
Merge Gateway $1 $3.2 31 Jul 2026
Nebius Token Factory $1 $3.2 31 Jul 2026
OpenCode Zen $1 $3.2 31 Jul 2026
OrcaRouter $1 $3.2 31 Jul 2026
Poe $1 $3.2 31 Jul 2026
Together AI $1 $3.2 31 Jul 2026
Vercel $1 $3.2 31 Jul 2026
Z.ai $1 $3.2 31 Jul 2026
Cortecs $1.08 $3.44 31 Jul 2026
NanoGPT TEE/glm-5 $1.2 $3.5 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, programming 1,460.36 1,453–1,468
Arena Score in French 1,457.22 1,436–1,479
Arena Score, expert questions 1,456.83 1,444–1,469
Arena Score in English 1,456.77 1,451–1,463
Arena Score, multi-turn dialogue 1,456.52 1,447–1,466
Arena Score, hard prompts 1,451.58 1,446–1,457
Arena Score, long queries 1,447.52 1,441–1,454
Arena Score in Spanish 1,446.57 1,427–1,466
Arena Score, overall 1,445.17 1,441–1,450
Arena Score, mathematics 1,438.25 1,424–1,453
Arena Score, creative writing 1,436.16 1,427–1,445
Arena Score in Russian 1,435.63 1,425–1,447
Arena Score, web development 1,434.74 1,426–1,443
Arena Score, instruction following 1,427.44 1,421–1,434
GPQA Diamond — graduate-level questions 83.76 % · with a tuned harness Epoch evaluations
Mock AIME 2024–2025 — olympiad problems 79.98 % · with a tuned harness Epoch evaluations
SWE-bench Verified — fixing bugs in repositories 72.08 % · with a tuned harness Epoch evaluations
Terminal-Bench — working in the command line 52.4 % · with a tuned harness
WeirdML — unusual machine learning tasks 48.17 % WeirdML Leaderboard
ARC-AGI — generalising to unseen patterns 44.67 % · with a tuned harness
SimpleBench — trick questions 43.84 % SimpleBench Leaderboard
APEX-Agents 17.2 %
Chess puzzles 5.3 % Epoch evaluations
ARC-AGI-2 4.86 %