Zhipu AI model product
GLM 5.3 Flash
3-Flash is a natively multimodal 320B-total/18B-active MoE model from Zhipu AI with a 1M-token context window.
Updated Sep 8, 2026. Default version: GLM-5.3-Flash
LLMBoard Score84.3GLM-5.3-Flash
Coverage20%14 benchmark families
Context window1MTokens
Official input price$0.075Zhipu AI API
GLM 5.3 Flash Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
GLM 5.3 Flash Benchmark Results
Benchmark scores for GLM-5.3-Flash.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkChartography | Score78.00% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkMMVU | Score80.50% | Rank01 | Participants5 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkMVBench | Score77.80% | Rank01 | Participants18 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkToolathlon | Score78.40% | Rank01 | Participants41 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkArtificial Analysis | Score57.00% | Rank02 | Participants8 | Percentile85.71% | EvidenceC | Evaluated |
| BenchmarkAutomationBench | Score48.80% | Rank02 | Participants18 | Percentile94.12% | EvidenceC | Evaluated |
| BenchmarkGDPval-AA | Score1,773.00 points | Rank02 | Participants9 | Percentile87.50% | EvidenceC | Evaluated |
| BenchmarkLM Arena Agent Task Outcome Explicit | Score14.14% | Rank04 | Participants49 | Percentile93.75% | EvidenceA | Evaluated |
| BenchmarkNL2Repo | Score56.30% | Rank05 | Participants22 | Percentile80.95% | EvidenceC | Evaluated |
| BenchmarkOfficeQA Pro | Score62.40% | Rank06 | Participants9 | Percentile37.50% | EvidenceC | Evaluated |
| BenchmarkBabyVision | Score53.40% | Rank08 | Participants11 | Percentile30.00% | EvidenceC | Evaluated |
| BenchmarkCharXiv-R | Score89.40% | Rank09 | Participants55 | Percentile85.19% | EvidenceC | Evaluated |
| BenchmarkLM Arena Agent Tool Hallucination | Score0.78% | Rank10 | Participants49 | Percentile81.25% | EvidenceA | Evaluated |
| BenchmarkLM Arena Vision | Score1,296.37 rating | Rank12 | Participants107 | Percentile89.62% | EvidenceA | Evaluated |
| BenchmarkAgents' Last Exam | Score26.30% | Rank13 | Participants16 | Percentile20.00% | EvidenceC | Evaluated |
| BenchmarkTerminal-Bench 2.1 | Score84.30% | Rank13 | Participants35 | Percentile64.71% | EvidenceC | Evaluated |
| BenchmarkLM Arena Webdev | Score1,604.97 rating | Rank13 | Participants97 | Percentile87.50% | EvidenceA | Evaluated |
| BenchmarkDeepSWE 1.1 | Score63.40% | Rank14 | Participants31 | Percentile56.67% | EvidenceC | Evaluated |
| BenchmarkHumanity's Last Exam | Score55.30% | Rank14 | Participants103 | Percentile87.25% | EvidenceC | Evaluated |
| BenchmarkLM Arena Agent Praise Complaint | Score5.45% | Rank16 | Participants49 | Percentile68.75% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text Factuality | Score1,471.34 rating | Rank17 | Participants121 | Percentile86.67% | EvidenceA | Evaluated |
| BenchmarkLM Arena Agent Leaderboard | Score3.67% | Rank19 | Participants49 | Percentile62.50% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text | Score1,470.84 rating | Rank20 | Participants210 | Percentile90.91% | EvidenceA | Evaluated |
| BenchmarkLM Arena Vision Style Control | Score1,272.63 rating | Rank21 | Participants107 | Percentile81.13% | EvidenceA | Evaluated |
| BenchmarkLM Arena Agent Steerability | Score0.23% | Rank22 | Participants49 | Percentile56.25% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text Style Control | Score1,474.16 rating | Rank23 | Participants210 | Percentile89.47% | EvidenceA | Evaluated |
| BenchmarkLM Arena Agent Bash Recovery Steps | Score-2.24% | Rank34 | Participants49 | Percentile31.25% | EvidenceA | Evaluated |
GLM 5.3 Flash Arena Results
Preference and agent-evaluation results for the default version.
Sort by
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenaagent task outcome explicit | Categoryoverall | Rank04 | Rating / score0.14 | VotesN/A | Observations13.5K | Result date |
| Arenatext | Categorycoding | Rank05 | Rating / score1,516.19 | Votes1,274 | ObservationsN/A | Result date |
| Arenatext factuality | Categorycoding | Rank05 | Rating / score1,536.12 | Votes1,243 | ObservationsN/A | Result date |
| Arenatext | Categoryexpert | Rank06 | Rating / score1,518.91 | Votes480 | ObservationsN/A | Result date |
| Arenatext style control | Categorycoding | Rank07 | Rating / score1,534.08 | Votes1,274 | ObservationsN/A | Result date |
| Arenavision | Categorydiagram | Rank08 | Rating / score1,320.60 | Votes314 | ObservationsN/A | Result date |
| Arenawebdev | Categorywebdev-react | Rank09 | Rating / score1,635.16 | Votes1,291 | ObservationsN/A | Result date |
| Arenaagent tool hallucination | Categoryoverall | Rank10 | Rating / score0.01 | VotesN/A | Observations1.8M | Result date |
| Arenatext | Categoryindustry software and it services | Rank11 | Rating / score1,499.95 | Votes1,807 | ObservationsN/A | Result date |
| Arenatext style control | Categorychinese | Rank11 | Rating / score1,531.54 | Votes270 | ObservationsN/A | Result date |
| Arenatext style control | Categoryexpert | Rank11 | Rating / score1,517.16 | Votes480 | ObservationsN/A | Result date |
| Arenavision | Categoryoverall | Rank12 | Rating / score1,296.37 | Votes1,389 | ObservationsN/A | Result date |
| Arenatext | Categorychinese | Rank13 | Rating / score1,527.63 | Votes270 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry software and it services | Rank13 | Rating / score1,513.77 | Votes1,788 | ObservationsN/A | Result date |
| Arenawebdev | Categoryoverall | Rank13 | Rating / score1,604.97 | Votes1,961 | ObservationsN/A | Result date |
| Arenawebdev | Categorywebdev | Rank13 | Rating / score1,604.97 | Votes1,961 | ObservationsN/A | Result date |
| Arenavision | Categoryenglish | Rank14 | Rating / score1,294.44 | Votes441 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts | Rank15 | Rating / score1,486.20 | Votes3,092 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry software and it services | Rank15 | Rating / score1,513.97 | Votes1,807 | ObservationsN/A | Result date |
| Arenavision | Categoryocr | Rank15 | Rating / score1,301.34 | Votes968 | ObservationsN/A | Result date |
| Arenaagent praise complaint | Categoryoverall | Rank16 | Rating / score0.05 | VotesN/A | Observations5.4K | Result date |
| Arenatext | Categoryhard prompts english | Rank16 | Rating / score1,483.31 | Votes1,147 | ObservationsN/A | Result date |
| Arenatext factuality | Categorynon english | Rank16 | Rating / score1,463.14 | Votes2,909 | ObservationsN/A | Result date |
| Arenavision style control | Categorydiagram | Rank16 | Rating / score1,308.37 | Votes314 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry mathematical | Rank17 | Rating / score1,490.60 | Votes225 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryinstruction following | Rank17 | Rating / score1,473.45 | Votes1,629 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryoverall | Rank17 | Rating / score1,471.34 | Votes4,669 | ObservationsN/A | Result date |
| Arenatext | Categoryinstruction following | Rank18 | Rating / score1,469.79 | Votes1,660 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryhard prompts | Rank18 | Rating / score1,496.13 | Votes3,091 | ObservationsN/A | Result date |
| Arenaagent | Categoryoverall | Rank19 | Rating / score0.04 | VotesN/A | Observations1.9M | Result date |
GLM 5.3 Flash Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.075 input, $0.25 output per 1M
- Official provider
- Zhipu AI
- Lowest third-party
- From $0.015 input, $0.05 output per 1M via Merge Gateway
- Tracked offerings
- 46
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderKenari | Provider model IDglm-5-3-flash | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderZhipu AI Coding Plan | Provider model IDglm-5.3-flash | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderZ.AI Coding Plan | Provider model IDglm-5.3-flash | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderNaN | Provider model IDglm5.3-flash | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderSCNet Token Plan | Provider model IDGLM-5.3-Flash | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderMerge Gateway | Provider model IDzai/glm-5.3-flash | Regionglobal | Input / 1M$0.015 | Output / 1M$0.05 | Context1M | Updated |
| ProviderVancine | Provider model IDglm-5.3-flash | Regionglobal | Input / 1M$0.06 | Output / 1M$0.20 | Context1M | Updated |
| ProviderCrofAI | Provider model IDglm-5.3-flash | Regionglobal | Input / 1M$0.07 | Output / 1M$0.22 | Context1M | Updated |
| ProviderTokenGo | Provider model IDz-ai/glm-5.3-flash | Regionglobal | Input / 1M$0.075 | Output / 1M$0.025 | Context1M | Updated |
| ProviderOrcaRouter | Provider model IDz-ai/glm-5.3-flash | Regionglobal | Input / 1M$0.075 | Output / 1M$0.25 | Context1M | Updated |
| ProviderNanoGPT | Provider model IDz-ai/glm-5.3-flash | Regionglobal | Input / 1M$0.075 | Output / 1M$0.25 | Context1M | Updated |
| ProviderOpenRouter | Provider model IDz-ai/glm-5.3-flash | Regionglobal | Input / 1M$0.075 | Output / 1M$0.25 | Context1.3M | Updated |
| ProviderCrossModel | Provider model IDz-ai/glm-5.3-flash | Regionglobal | Input / 1M$0.075 | Output / 1M$0.25 | Context1M | Updated |
| ProviderZ.AI | Provider model IDglm-5.3-flash | Regionglobal | Input / 1M$0.075 | Output / 1M$0.25 | Context1M | Updated |
| ProviderOpenCode Go | Provider model IDglm-5.3-flash | Regionglobal | Input / 1M$0.075 | Output / 1M$0.25 | Context1M | Updated |
| ProviderOfox | Provider model IDz-ai/glm-5.3-flash | Regionglobal | Input / 1M$0.075 | Output / 1M$0.25 | Context1M | Updated |
| ProviderZhipu AI | Provider model IDglm-5.3-flash | Regionglobal | Input / 1M$0.075 | Output / 1M$0.25 | Context1M | Updated |
| ProviderEmpirioLabs AI | Provider model IDglm-5-3-flash | Regionglobal | Input / 1M$0.075 | Output / 1M$0.25 | Context1M | Updated |
| ProviderDevPass (LLM Gateway) | Provider model IDglm-5.3-flash | Regionglobal | Input / 1M$0.10 | Output / 1M$0.25 | Context1M | Updated |
| ProviderRunInfra | Provider model IDzai-org/GLM-5.3-Flash | Regionglobal | Input / 1M$0.10 | Output / 1M$0.40 | Context1M | Updated |
| ProviderAIHubMix | Provider model IDglm-5.3-flash | Regionglobal | Input / 1M$0.1127 | Output / 1M$0.3944 | Context1M | Updated |
| ProviderCloudflare Workers AI | Provider model ID@cf/zai-org/glm-5.3-flash | Regionglobal | Input / 1M$0.15 | Output / 1M$0.50 | Context1.3M | Updated |
| ProviderSynthetic | Provider model IDhf:zai-org/GLM-5.3-Flash | Regionglobal | Input / 1M$0.15 | Output / 1M$0.50 | Context524.3K | Updated |
| ProviderDeep Infra | Provider model IDzai-org/GLM-5.3-Flash | Regionglobal | Input / 1M$0.15 | Output / 1M$0.50 | Context1M | Updated |
| ProviderHugging Face | Provider model IDzai-org/GLM-5.3-Flash | Regionglobal | Input / 1M$0.15 | Output / 1M$0.50 | Context1M | Updated |
| ProviderVivgrid | Provider model IDglm-5.3-flash | Regionglobal | Input / 1M$0.15 | Output / 1M$0.50 | Context1M | Updated |
| ProviderNebius Token Factory | Provider model IDzai-org/GLM-5.3-Flash | Regionglobal | Input / 1M$0.15 | Output / 1M$0.50 | Context1M | Updated |
| ProviderDigitalOcean | Provider model IDglm-5.3-flash | Regionglobal | Input / 1M$0.15 | Output / 1M$0.50 | Context1M | Updated |
| ProviderTogether AI | Provider model IDzai-org/GLM-5.3-Flash | Regionglobal | Input / 1M$0.15 | Output / 1M$0.50 | Context1M | Updated |
| ProviderBaseten | Provider model IDzai-org/GLM-5.3-Flash | Regionglobal | Input / 1M$0.15 | Output / 1M$0.50 | Context1M | Updated |
GLM 5.3 Flash Runtime Performance
Provider-specific output speed and catalog latency for GLM-5.3-Flash. Runtime does not affect the capability score.
Sort by
Provider | Output Speed | Catalog Latency | Max Input | Max Output | Updated |
|---|
| ProviderFriendliAI | Output Speed74.28 tok/s | Catalog Latency1.57 s | Max Input1M | Max Output131.1K | Updated |
| ProviderDeepInfra | Output Speed38.15 tok/s | Catalog Latency2.64 s | Max Input1M | Max Output131.1K | Updated |
| ProviderZAI | Output Speed7.17 tok/s | Catalog Latency9.40 s | Max Input1M | Max Output131.1K | Updated |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
GLM 5.3 Flash Specifications
Technical details for the model's default version.
- Version
- GLM-5.3-Flash
- Released
- Aug 26, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 320B
- Context window
- 1M
- Max output
- 131.1K
- Inputs
- image, text, video
- Outputs
- text
- Open weights
- Yes
- License
- MIT
GLM 5.3 Flash Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionGLM-5.3-Flash | Released | LLMBoard84.28 | Parameters320B | Context1M | Max output131.1K | Open weightsYes | LicenseMIT |
Models similar to GLM 5.3 Flash
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
ZA
GLM 5.3
Zhipu AI
ZA
GLM 5.2
Zhipu AI
ZA
GLM 5.1
Zhipu AI
ZA
GLM 5
Zhipu AI
ZA
GLM 4.7
Zhipu AI
ZA
GLM 5V Turbo
Zhipu AI
What is GLM 5.3 Flash?
Key information about GLM 5.3 Flash and its available data.
AI's GLM-5 series, using hybrid sparse and linear attention with mHC. It has MIT-licensed open weights available on Hugging Face.
Data as of 2026-09-08.
FAQ
Common questions about GLM 5.3 Flash.
When was GLM 5.3 Flash released?
GLM 5.3 Flash's default version was released on Aug 26, 2026.
How much does GLM 5.3 Flash cost?
GLM 5.3 Flash's official API price is $0.075 per million input tokens and $0.25 per million output tokens via Zhipu AI. The lowest tracked third-party offer starts at $0.015 input and $0.05 output via Merge Gateway.
Who created GLM 5.3 Flash?
GLM 5.3 Flash was created by Zhipu AI.
What is the context window for GLM 5.3 Flash?
The default version has a 1M token context window.
Is GLM 5.3 Flash open weight?
Yes. The default version is marked as open weight under MIT.
How many API providers offer GLM 5.3 Flash?
46 provider offerings are linked to the default version.