Identity
- Status
- active
- Developer ID
glm-5.3-flash- Released
- Sep 2, 2026
- Version
- GLM-5.3-Flash
- License
- MIT
Z.aiGLM-5.3-Flash
GLM-5.3-Flash is a native multimodal model with 320B total and 18B active parameters, a 1M-token context window, and always-on reasoning.
- Context
- 1,000,000 tokens
- Inputs
- Text, Image, Video, Document
- Outputs
- Text
- Released
- Sep 2, 2026
Providers & Pricing
Estimate workload cost →| Provider | Input / 1M | Output / 1M | Cache read / 1M | Cache write / 1M | Observed | Evidence |
|---|---|---|---|---|---|---|
 DeepinfraAvailable | $0.15 | $0.50 | $0.030 | — | Sep 2, 2026 | Provider-reported |
Z.aiAvailable | $0.075 | $0.25 | $0.015 | — | Aug 29, 2026 | Provider-reported |
Benchmarks
All benchmark results →| GLM-5.3-FlashZ.ai | LiveBench2026-06-25 | 73.27 | 34,707 | 1,270 | glm-5.3-flashaverage per case | Recomputedwinner eligible | Third-party benchmark |
| GLM-5.3-FlashZ.ai | LMArena Text Arenatext-2026-09-01-011508720696 | 1,470.91 | — | 4,399 | glm-5.3-flash; 95% CI [1461.80526636, 1480.01283657]; votes 4399; rank 23unknown | Original sourcewinner eligible | Third-party benchmark |
| GLM-5.3-FlashZ.ai | LMArena Vision Arenavision-2026-08-27-011508720696 | 1,296.37 | — | 1,389 | glm-5.3-flash; 95% CI [1279.58251138, 1313.15015420]; votes 1389; rank 15unknown | Original sourcewinner eligible | Third-party benchmark |
At a Glance
Model Facts
Capacity
- Context
- 1,000,000 tokens
- Maximum output
- 131,072 tokens
- Total parameters
- 320,000,000,000
- Active parameters
- 18,000,000,000
- Knowledge cutoff
- Not reported
Interface and Access
- Inputs
- Text, Image, Video, Document
- Outputs
- Text
- API available
- Yes
- Open weights
- Yes
- Self-hostable
- Yes
- Reasoning
- Yes
- Vision
- Yes
- Tool calling
- Yes
- Capabilities
- agents, chat, computer-use, reasoning, structured outputs, tools, vision
This page represents one immutable developer model ID. Serving endpoints, pricing, regions, quantizations, and service tiers attach separately and never create duplicate model records.
Version Lineage
PredecessorUnknown
SuccessorsUnknown
Aliasesglm-5.3-flash, glm-5.3-flash, ox-alpha, zai-org/GLM-5.3-Flash
Explicit Unknowns
- Tokenizer
- Knowledge cutoff
Primary Evidence
Sources and Observation Date
GLM-5.3-Flash | Z.AI Developer DocumentZ.ai · observed Sep 2, 2026GLM-5.3-Flash API pricing | Z.AIZ.ai · observed Sep 2, 2026GLM-5.3-Flash: More Intelligence with Less ComputeZ.ai · observed Sep 2, 2026zai-org/glm-5.3-flash immutable Hub API metadataZ.ai · observed Sep 2, 2026Deepinfra catalogDeepinfra · observed Sep 2, 2026
Comparable Peers
Related Model Comparisons
| A | Pair | B | Context |
|---|---|---|---|
GLM-5.3-FlashZ.ai | vs |  DeepSeek-V4-Flash-Vision-ExpDeepSeek | cross-developer peersimage, text |
GLM-5.3-FlashZ.ai | vs | GLM-5V-TurboZ.ai | family variantsimage, text, video |
GLM-5.3-FlashZ.ai | vs |  Nova 2 LiteAmazon | cross-developer peersimage, text, video |
GLM-5.3-FlashZ.ai | vs |  Mistral Large 3Mistral AI | cross-developer peersimage, text |
GLM-5.3-FlashZ.ai | vs |  Command A VisionCohere | cross-developer peersimage, text |
GLM-5.3-FlashZ.ai | vs | GLM-OCRZ.ai | family variantsimage, text |
GLM-5.3-FlashZ.ai | vs |  Gemini 3.5 FlashGoogle DeepMind | cross-developer peerstext |
GLM-5.3-FlashZ.ai | vs | Gemini 3 FlashGoogle DeepMind | cross-developer peerstext |
GLM-5.3-FlashZ.ai | vs | GLM-5.3Z.ai | family variantstext |
GLM-5.3-FlashZ.ai | vs | GLM-5.2Z.ai | family variantstext |
GLM-5.3-FlashZ.ai | vs | DeepSeek-V4-FlashDeepSeek | cross-developer peerstext |
GLM-5.3-FlashZ.ai | vs | DeepSeek-V4-Flash-BaseDeepSeek | cross-developer peerstext |
Questions
GLM-5.3-Flash FAQs
What is GLM-5.3-Flash?+
GLM-5.3-Flash is a native multimodal model with 320B total and 18B active parameters, a 1M-token context window, and always-on reasoning. It is developed by Z.ai and its current sourced lifecycle status is active.
What inputs and outputs does GLM-5.3-Flash support?+
GLM-5.3-Flash's current record lists Text, Image, Video, and Document as input and Text as output.
How much context does GLM-5.3-Flash support?+
The current record lists a 1,000,000-token context window and a maximum output of 131,072 tokens.
Is GLM-5.3-Flash available through an API?+
GLM-5.3-Flash is marked as API-available. The directly sourced provider records currently include Deepinfra and Z.ai.
How large is GLM-5.3-Flash?+
The current record lists 320,000,000,000 total parameters and 18,000,000,000 active parameters. Its architecture is hybrid_sparse_linear_attention_moe.
What capabilities and tasks does GLM-5.3-Flash support?+
GLM-5.3-Flash's recorded capabilities are agents, chat, computer-use, reasoning, structured outputs, tools, and vision. Its supported tasks are coding, computer-use, document-question-answering, image, text, and video.
How much does GLM-5.3-Flash cost through an API?+
The lowest directly sourced prices currently attached to GLM-5.3-Flash are $0.075 per million input tokens through Z.ai and $0.25 per million output tokens through Z.ai. Prices are provider-specific and should be checked against each cited observation date.
Can GLM-5.3-Flash be self-hosted?+
GLM-5.3-Flash is an open-weight model. Self-hosting is supported, and the recorded license is MIT.
Does GLM-5.3-Flash have related versions or aliases?+
No predecessor is recorded. No successor is recorded. Recorded aliases are glm-5.3-flash, ox-alpha, and zai-org/GLM-5.3-Flash.
Which models can GLM-5.3-Flash be compared with?+
The current compatible peer set includes DeepSeek-V4-Flash-Vision-Exp, GLM-5V-Turbo, Nova 2 Lite, Mistral Large 3, and Command A Vision. Compatibility requires shared sourced modalities and tasks; it does not imply that one model is better. Browse model comparisons →