Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models762Deals30d246Policy30d61, 12 pendingSources1,061

GLM-5.3-Flash

Z.aiReleased Aug 26, 2026Current

Score basis

GLM-5.3-Flash is capable in long context, coding, reasoning, and agentic tasks; and behind the leaders in factuality and math. Too few results yet to rate safety, multimodal tasks, multilingual tasks, or instruction following.

Capability profile

Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.

Capable4 capabilities
  1. Long Context−11.2%1 of 3
  2. Coding−17.1%9 of 10
  3. Reasoning−18.6%5 of 6
  4. Agentic−21.3%6 of 7
Limited2 capabilities
  1. Factuality−27.1%3 of 4
  2. Math−49.8%3 of 5
Not rated4 capabilities

Too few results yet to rate Safety, Multimodal, Multilingual or Instruction Following.

Price

$0.15input$0.50outputper million tokens

From Artificial Analysis · 7 providers tracked · All prices

Cheaper than 77% of 329 priced models · 3:1 input-to-output blend, log scale

Evidence

76results on65benchmarks

  • 20 independently verified
  • 16 aggregator
  • 12 vendor-reported
  • 28 cross-referenced

From 13 sources · latest Oct 17, 2026 · How verification works

API features

Tool callingStructured outputsReasoning

As listed by OpenRouter

Research

9 papers reference GLM-5.3-Flash

GLM-5.3-Flash benchmark results

76 results on 65 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.

Long Context

CapableFull long context ranking

11.2% behind the leader1 of 3 ranked benchmarks measured

Coding

CapableFull coding ranking

17.1% behind the leader9 of 10 ranked benchmarks measured

Show 2 more coding resultsHide 2 coding results

Reasoning

CapableFull reasoning ranking

18.6% behind the leader5 of 6 ranked benchmarks measured

Show 4 more reasoning resultsHide 4 reasoning results

Agentic

CapableFull agentic ranking

21.3% behind the leader6 of 7 ranked benchmarks measured

Show 1 more agentic resultHide 1 agentic result

Factuality

LimitedFull factuality ranking

27.1% behind the leader3 of 4 ranked benchmarks measured

Math

LimitedFull math ranking

49.8% behind the leader3 of 5 ranked benchmarks measured

Multimodal

Not enough dataFull multimodal ranking

0 of 6 ranked benchmarks measured

Instruction Following

Not enough dataFull instruction following ranking

0 of 3 ranked benchmarks measured

More results

Benchmarks outside the capability baskets. They are not ranked against a leader.

Show 32 more resultsHide 32 results

GLM-5.3-Flash: common questions

Who makes GLM-5.3-Flash?

GLM-5.3-Flash is made by Z.ai.

When was GLM-5.3-Flash released?

GLM-5.3-Flash was released on Aug 26, 2026, according to Artificial Analysis.

What is GLM-5.3-Flash good at?

GLM-5.3-Flash is capable in long context, coding, reasoning, and agentic tasks; and behind the leaders in factuality and math. Too few results yet to rate safety, multimodal tasks, multilingual tasks, or instruction following.

How much does GLM-5.3-Flash cost?

GLM-5.3-Flash costs $0.15 per million input tokens and $0.50 per million output tokens, according to Artificial Analysis. We track its price at 7 providers. At a mix of three input tokens to one output token, it is cheaper than 77% of the 329 priced models we track.

How many benchmarks has GLM-5.3-Flash been tested on?

We track 76 results for GLM-5.3-Flash on 65 benchmarks from 13 sources, 20 of them independently verified. The latest was recorded on Oct 17, 2026.

Which API features does GLM-5.3-Flash support?

OpenRouter lists tool calling, structured outputs, and reasoning for GLM-5.3-Flash.

About this record

Where GLM-5.3-Flash's numbers come from, and every name it appears under.

Tracked since
Aug 26, 2026
Newest source mention
Oct 9, 2026

Where the results come from

Verification: 76 scores · 20 independently verified · 16 aggregator-attributed · 28 vendor cross-reference · 12 vendor-reported. How these tiers are assigned

From 13 sources on 8 sites. Hugging Face supplies 28 of them; the 20 independently verified results come from 6 sites. Bars are coloured by trust tier.

  • huggingface.co28
  • artificialanalysis.ai17
  • api.llm-stats.com12
  • livebench.ai7
  • arcprize.org6
  • epoch.ai3
  • datasets-server.huggingface.co2
  • lmarena.ai1

Also known as

How our sources name GLM-5.3-Flash at each reasoning setting.

SettingShort formAPI id
low—glm-5.3-flash (low)
high—glm-5.3-flash (high)
maxglm 5.3 flash maxglm-5.3-flash (max)
Also listed asglm-5-3-flashglm-5.3-flash-bf16glm-5.3-flash-ggufglm-5.3-flash:batchglm‑5.3‑flashglm5.3 flashswift glm 5.3 flashunsloth/glm-5.3-flash-ggufzai-org/glm-5.3-flash