GLM 5.3

Z.ai · On OpenRouter since August 18, 2026 · Pinned version id: z-ai/glm-5.3-20260816

Price

$2.15

mixed / 1M tokens

Mixed price per 1M tokens: 3 input tokens to 1 output token.

in $1.40 · out $4.40 · Cached input $0.26

Models that think before they answer can use many more output tokens, so your real cost can be higher than the figure shown.

US dollars per million tokens, converted to your currency at the NBP rate.

Tasks, with the evidence

Each line names the measurement it rests on. A task that is not listed has no measurement or declared feature we can show.

Writing code Top 25%

Measured by a third party

Code: 1,275 · Rank 23 of 122, among scored models · Design Arena, read November 21, 2026

Cost of this job: $1.01

Hard questions and reasoning

Measured by a third party

Intelligence Index: 31.7 · Rank 27 of 66, among scored models · Artificial Analysis, read November 21, 2026

Cost of this job: $0.39

Models that score at least as high for less in this scenario: 6

Websites and apps Top 25%

Measured by a third party

Websites: 1,273 · Rank 26 of 127, among scored models · Top 25% · Design Arena, read November 21, 2026

UI components: 1,277 · Rank 27 of 118, among scored models · Top 25% · Design Arena, read November 21, 2026

Web apps (agent): 1,270 · Rank 8 of 47, among scored models · Top 25% · Design Arena, read November 21, 2026

Mobile apps (agent): 1,183 · Rank 25 of 50, among scored models · Design Arena, read November 21, 2026

Cost of this job: $0.46

Models that score at least as high for less in this scenario: 9

Games and 3D Top 25%

Measured by a third party

Game development: 1,276 · Rank 23 of 121, among scored models · Top 25% · Design Arena, read November 21, 2026

3D scenes: 1,278 · Rank 22 of 114, among scored models · Top 25% · Design Arena, read November 21, 2026

Cost of this job: $0.46

Models that score at least as high for less in this scenario: 7

Slides and spreadsheets

Measured by a third party

PowerPoint files (agent): 1,217 · Rank 9 of 29, among scored models · Design Arena, read November 21, 2026

HTML slides (agent): 1,189 · Rank 2 of 25, among scored models · Design Arena, read November 21, 2026

Cost of this job: $0.46

Charts and graphics

Measured by a third party

Data visualization: 1,245 · Rank 37 of 112, among scored models · Design Arena, read November 21, 2026

SVG graphics: 1,238 · Rank 21 of 84, among scored models · Top 25% · Design Arena, read November 21, 2026

ASCII art: 1,182 · Rank 31 of 68, among scored models · Design Arena, read November 21, 2026

Cost of this job: $0.46

Models that score at least as high for less in this scenario: 16

Long documents

Declared, not measured

Context window: 1M tokens as declared. It says how much fits, not how well the model uses it.

Cost of this job: $1.73

The size of each job is our assumption; the prices are real.

  • Writing code: Requests: 40. Input tokens per request: 30,000. Output tokens per request: 1,500. Share of input served from cache: 70%.
  • Hard questions and reasoning: Requests: 10. Input tokens per request: 3,000. Output tokens per request: 8,000. Share of input served from cache: 0%.
  • Websites and apps; Games and 3D; Slides and spreadsheets; Charts and graphics: Requests: 8. Input tokens per request: 12,000. Output tokens per request: 10,000. Share of input served from cache: 30%.
  • Long documents: Requests: 3. Input tokens per request: 400,000. Output tokens per request: 4,000. Share of input served from cache: 0%.

Thinking is billed as output tokens; a model that thinks longer costs more than this estimate. Ranks are among all models we list, including older ones. The share is rounded up to whole models, so a rank slightly past the exact share still counts. Models with equal scores share the better rank.

Cheaper models with a similar benchmark score

Rule: an Intelligence Index of at least 28.7, a mixed price at least 25% lower, at least the same input types, and tool support. Lower mixed price (3 input tokens to 1 output token); check the flags and your own prompt sizes.

  1. Ling 3.1 Flash inclusionAI 96% cheaper $0.09 · Index 29.6 (this model: 31.7) Differences to check: No structured output; Lower max output Compare
  2. Claude Haiku 5.5 Anthropic 91% cheaper $0.20 · Index 36.5 (this model: 31.7) Differences to check: Lower max output; Price rises for long prompts Compare
  3. Muse Ember Meta 91% cheaper $0.20 · Index 36.0 (this model: 31.7) Compare
  4. MiniMax M3.1 Flash MiniMax 88% cheaper $0.2625 · Index 29.1 (this model: 31.7) Compare
  5. Inkling 2 Small Thinking Machines 78% cheaper $0.475 · Index 30.6 (this model: 31.7) Compare
Show more: 4
  1. DeepSeek V4.1 Flash 1017 DeepSeek 76% cheaper $0.525 · Index 32.6 (this model: 31.7) Differences to check: Lower max output Compare
  2. MiMo-V2.6-Pro Xiaomi 75% cheaper $0.5438 · Index 29.9 (this model: 31.7) Differences to check: Lower max output Compare
  3. Hy4 Tencent 71% cheaper $0.6254 · Index 34.6 (this model: 31.7) Differences to check: Lower max output Compare
  4. Gemini 3.8 Flash Google 30% cheaper $1.50 · Index 30.8 (this model: 31.7) Differences to check: Lower max output Compare

Price by provider

The price used in the catalog is the vendor's own; when the vendor has no API of its own, it is the cheapest regular host.

ProviderInputOutputCached inputMixedNotes
Wafer $0.0699$1.40 $0.0689 $0.4024 81% cheaper US · fp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Wafer $0.179$1.40 $0.178 $0.4842 77% cheaper fp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Decart $0.56$1.81 $0.10 $0.8725 59% cheaper fp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Inference.net $0.084$3.60 $0.079 $0.963 55% cheaper fp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Inference.net $0.084$3.60 $0.079 $0.963 55% cheaper US · fp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Novita $0.63$1.98 $0.117 $0.9675 55% cheaper fp8 · promo −55%
SiliconFlow $0.70$1.90 $0.13 $1.00 53% cheaper fp8
DeepInfra $0.5625$2.50 $0.125 $1.05 51% cheaper fp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality. · promo −38%
Show all providers: 43
  • Cloudflare: $0.70 / $2.20 · $1.08
  • Doubleword: $0.70 / $2.20 · $1.08
  • GMICloud: $0.70 / $2.20 · $1.08
  • Tenstorrent: $0.60 / $2.50 · $1.08
  • Relace: $0.0703 / $4.38 · $1.15
  • AkashML: $0.19 / $4.40 · $1.24
  • Cohere: $0.50 / $3.50 · $1.25
  • CoreWeave: $0.75 / $3.00 · $1.31
  • Sail Research: $0.75 / $3.00 · $1.31
  • DigitalOcean: $0.91 / $2.86 · $1.40
  • Featherless: $0.98 / $3.08 · $1.51
  • Inceptron: $1.05 / $3.50 · $1.66
  • Alibaba: $1.19 / $3.74 · $1.83
  • Sail Research: $1.05 / $4.20 · $1.84
  • Friendli: $1.26 / $3.96 · $1.94
  • Morph: $1.10 / $5.12 · $2.10
  • Z.ai: $1.40 / $4.40 · $2.15
  • AtlasCloud: $1.40 / $4.40 · $2.15
  • Baidu: $1.40 / $4.40 · $2.15
  • BaseTen: $1.40 / $4.40 · $2.15
  • BaseTen: $1.40 / $4.40 · $2.15
  • Crusoe: $1.40 / $4.40 · $2.15
  • Doubleword: $1.40 / $4.40 · $2.15
  • Fireworks: $1.40 / $4.40 · $2.15
  • Mistral: $1.40 / $4.40 · $2.15
  • Modal: $1.40 / $4.40 · $2.15
  • Nebius: $1.40 / $4.40 · $2.15
  • Parasail: $1.40 / $4.40 · $2.15
  • SambaNova: $1.40 / $4.40 · $2.15
  • Together: $1.40 / $4.40 · $2.15
  • Venice: $1.40 / $4.40 · $2.15
  • Makora: $1.50 / $4.40 · $2.23
  • Mistral: $1.54 / $4.84 · $2.37
  • Amazon Bedrock: $1.68 / $5.28 · $2.58
  • ModelRun: $2.10 / $6.60 · $3.23

Other pricing modes

Specifications

Context window
1M
Max output
944k
Accepts
text
Answers with
text
Tools
Yes
Structured output
Yes
Reasoning
Always on
Thinking effort levels: low, high, and max
Thinking is on by default
Weights on Hugging Face
Hugging Face A Hugging Face repository is linked. We did not check the licence; read it before you build on it.
Knowledge cutoff
No data
On OpenRouter since
August 18, 2026
Price check
one source

As listed by OpenRouter. "Not listed" means OpenRouter does not list it for this model.

Benchmark scores

Scores come from third parties and are relayed by OpenRouter. We do not run the tests. Higher is better.

SourceScoreRank among scored
Intelligence Index
Artificial Analysis, read November 21, 2026
31.7Rank 27 of 66, among scored models
Websites
Design Arena, read November 21, 2026
1,273Rank 26 of 127, among scored models Top 25%
UI components
Design Arena, read November 21, 2026
1,277Rank 27 of 118, among scored models Top 25%
Game development
Design Arena, read November 21, 2026
1,276Rank 23 of 121, among scored models Top 25%
3D scenes
Design Arena, read November 21, 2026
1,278Rank 22 of 114, among scored models Top 25%
Data visualization
Design Arena, read November 21, 2026
1,245Rank 37 of 112, among scored models
SVG graphics
Design Arena, read November 21, 2026
1,238Rank 21 of 84, among scored models Top 25%
ASCII art
Design Arena, read November 21, 2026
1,182Rank 31 of 68, among scored models
Code
Design Arena, read November 21, 2026
1,275Rank 23 of 122, among scored models Top 25%
Web apps (agent)
Design Arena, read November 21, 2026
1,270Rank 8 of 47, among scored models Top 25%
Mobile apps (agent)
Design Arena, read November 21, 2026
1,183Rank 25 of 50, among scored models
PowerPoint files (agent)
Design Arena, read November 21, 2026
1,217Rank 9 of 29, among scored models
HTML slides (agent)
Design Arena, read November 21, 2026
1,189Rank 2 of 25, among scored models
agents/pdf
Design Arena, read November 21, 2026
1,279Rating 1,279; rank 5 at the source. Too few scored models for a tier.

Intelligence, Coding and Agentic indices: Artificial Analysis (artificialanalysis.ai). Design categories: Design Arena (designarena.ai). Both are relayed by the OpenRouter public model list and were read on November 21, 2026. We compute ranks, tiers and costs from these numbers; the sources have not reviewed them. Scores belong to the model, not to the route it is served on.

Artificial Analysis · Design Arena · Relayed by OpenRouter. opens in a new tab

Sources and dates

Prices: OpenRouter routes and a second price table, checked November 21, 2026. Specifications: OpenRouter, November 21, 2026. Exchange rate: NBP, November 20, 2026. Benchmarks: Artificial Analysis and Design Arena via OpenRouter.