Skip to content
  • Models
  • Rankings
  • Ori
Sign Up
Sign Up
OpenRouterOpenRouter
© 2026 OpenRouter, Inc

Product

  • Chat
  • Rankings
  • Benchmarks
  • Apps
  • Discover
  • Models
  • Collections
  • Providers
  • Pricing
  • Business
  • Enterprise
  • Labs

Company

  • About
  • Blog
  • Careers
    Hiring
  • Privacy
  • Terms of Service
  • Trust Center
  • Support
  • Works With OR
  • Data
  • Brand

Developer

  • Documentation
  • API Reference
  • Developer Platform
  • Status

Connect

  • Discord
  • GitHub
  • LinkedIn
  • X
  • YouTube
Favicon for z-ai

Z.ai: GLM 5.3

z-ai/glm-5.3

Model weights
Compare

GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks. It supports text input and output with a 1M-token context window, and improves on GLM-5.2 in coding and in the balance between performance and token efficiency.

Reasoning is always on and cannot be disabled. Reasoning efforts low, high, and max are supported; max is the default.

Modalities

In / Out Price

44% off

$0.7826 / $2.46per 1M

Context

1.3M

Released

Aug 18, 2026

Compare
BenchmarksProvidersPricingPerformanceUptimeAppsActivityFAQExplore

Benchmarks

Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.

Benchmark score summary for Z.ai: GLM 5.3 (Artificial Analysis and Design Arena)
SourceBenchmarkScore
Artificial AnalysisGLM-5.3 (max) Intelligence Index44.8
Artificial AnalysisGLM-5.3 (max) Coding Index74.8
Artificial AnalysisGLM-5.3 (max) Agentic Index53.1
Artificial AnalysisGLM-5.3 (max) GPQA Diamond91.7%
Artificial AnalysisGLM-5.3 (max) HLE42.3%
Artificial AnalysisGLM-5.3 (max) AA-LCR79.7%
Artificial AnalysisGLM-5.3 (max) GDPval-AA57.3%
Artificial AnalysisGLM-5.3 (max) CritPt19.1%
Artificial AnalysisGLM-5.3 (max) SciCode59.0%
Artificial AnalysisGLM-5.3 (max) AA-Omniscience Accuracy33.9%
Artificial AnalysisGLM-5.3 (max) AA-Omniscience Non-Hallucination Rate70.4%
Design ArenaGLM-5.3 Models Arena 3D Elo1390
Design ArenaGLM-5.3 Models Arena Asciiart Elo1243
Design ArenaGLM-5.3 Models Arena Code Categories Elo1327
Design ArenaGLM-5.3 Models Arena Data Visualization Elo1270
Design ArenaGLM-5.3 Models Arena Game Development Elo1374
Design ArenaGLM-5.3 Models Arena SVG Elo1318
Design ArenaGLM-5.3 Models Arena UI Component Elo1341
Design ArenaGLM-5.3 Models Arena Website Elo1312
Design ArenaGLM-5.3 Agents Arena Htmlslides Elo1189
Design ArenaGLM-5.3 Agents Arena Mobile Apps Elo1223
Design ArenaGLM-5.3 Agents Arena Python-Pptxslides Elo1258

Providers

Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick — Balanced (price + speed), Nitro (fastest), Floor (cheapest), or Exacto (highest tool-calling accuracy).

Pricing

The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.

Performance

Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).

Uptime

Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.

Apps

Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.

Activity

Token volume and request traffic to this model over time.

Quick Start

Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.

Explore more models

AI Model RankingsRanking
35% off
$1.19$0.7735$3.74$2.4310$0.1955$0.12711.09s55 tps
100.00%
44% off
$1.40$0.7826$4.40$2.4600$0.26$0.14532.96s42 tps
100.00%
33% off
$1.17$0.7839$3.96$2.6530$0.234$0.15681.95s60 tps
100.00%
38% off
$1.30$0.8060$4.40$2.7280$0.26$0.16121.60s121 tps
99.44%
40% off
$1.40$0.8400$4.40$2.6400$0.26$0.15603.73s33 tps
100.00%
$0.9000$3.0000$0.15001.42s86 tps
100.00%
25% off
$1.20$0.9000$4.00$3.0000$0.20$0.15001.41s74 tps
100.00%
$0.9100$2.8600$0.16900.97s51 tps
100.00%
$1.0056$3.4034$0.17790.99s78 tps
100.00%
$1.0250$3.2900$0.19001.39s36 tps
100.00%
25% off
$1.40$1.0500$4.40$3.3000$0.26$0.19501.92s54 tps
100.00%
$1.0500$4.2000$0.19000.95s85 tps
98.65%
20% off
$1.40$1.1200$4.40$3.5200$0.26$0.20801.11s55 tps
100.00%
$1.1900$3.7400$0.23801.26s42 tps
100.00%
15% off
$1.40$1.1900$4.40$3.7400$0.23$0.19551.20s97 tps
100.00%
$1.2070$3.8710$0.22500.76s27 tps
100.00%
10% off
$1.40$1.2600$4.40$3.9600$0.26$0.23400.49s134 tps
100.00%
$1.3000$4.4000$0.26000.65s118 tps
100.00%
$1.4000$4.4000$0.26002.22s45 tps
100.00%
$1.4000$4.4000$0.14001.85s53 tps
100.00%
$1.4000$4.4000$0.14001.14s131 tps
100.00%
$1.4000$4.4000$0.26001.12s92 tps
99.52%
$1.4000$4.4000$0.26000.77s124 tps
100.00%
20% off
$1.75$1.4000$5.50$4.4000$0.325$0.26001.95s107 tps
100.00%
$1.4000$4.4000$0.26000.40s146 tps
100.00%
$1.4000$4.4000$0.26000.87s128 tps
100.00%
$1.4000$4.4000$0.26001.03s127 tps
100.00%
$1.4000$4.4000$0.14001.55s68 tps
100.00%
$1.4000$4.4000$0.26001.42s63 tps
100.00%
$1.4000$4.4000$0.26003.77s9 tps
100.00%
$1.4000$4.4000$0.26006.47s24 tps
--
$1.4000$4.4000$0.26005.46s25 tps
100.00%
$2.1000$6.6000$0.21003.18s47 tps
100.00%
$2.1000$6.6000$0.21003.77s52 tps
100.00%

Throughput

146tok/s

P50, best across providers

Latency

0.40s

P50, best provider

AutoExacto Benchmarks
GPQA DiamondTAU-BenchBaidu Qianfan86.5%--Wafer91.8%78.6%Baseten (fp8) (US)83.5%--Baseten (fp8)86.4%79.2%Z.ai85.8%78.2%Decart86.1%77.6%Baseten (fp4) (US)86.4%76.7%Baseten (fp4)86.1%77.0%Baidu Qianfan84.9%76.0%
+27 more providers
Uptime (3d)The model was reachable. Request routed to a provider.

100.00%

Availability (3d)The model returned inference from any provider. Errors and empty responses count against it.

99.81%

Availability over the last 3 days

Last 72 hours
Availability 99.81%
3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability
99.77%
Without Routing
88.57%

When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and customization options.

Frequently asked questions

GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks. It supports text input and output with a 1M-token context window, and improves on GLM-5.2 in coding and in the balance between performance and token efficiency. Reasoning is always on and cannot be disabled. Reasoning efforts low, high, and max are supported; max is the default.

GLM 5.3 costs $0.7826/M input tokens and $2.46/M output tokens, with separate rates for Cache Read at $0.1453/M tokens.

GLM 5.3 has a 1,310,720 token context window. It supports up to 131,072 completion tokens.

Yes. GLM 5.3 accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.

GLM 5.3 is served by 30 providers on OpenRouter: Morph, NovitaAI, Reka AI, io.net, Phala, inference.net, DeepInfra, DigitalOcean and 22 more. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.

GLM 5.3 was released on August 18, 2026.