GLM 5.3
Z.ai (Zhipu AI)huggingface.co/zai-org/GLM-5.3
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...
Updated 4 h ago · first seen 11 Sept 2026
model_01M294AJ5462DT6GQKJAGHX7FP
Overview
Identity
Identity block not returned by the API for this entity.
Openness
Openness not classified yet — no sourced evidence to place this model in the ontology.
Key facts
- Release date
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 6 h agomedium
- Model card
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 11 h agomedium
- Openrouter id
Source:OpenRouter public model & pricing listingT2observed 11 h agomedium
Architecture
- Architecture
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 10 h agomedium
- Model type
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 10 h agomedium
- Parameters
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 11 h agomedium
- Weights dtype
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 10 h agomedium
- File size
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 10 h agomedium
- Library name
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 10 h agomedium
- Pipeline tag
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 11 h agomedium
- Hugging Face repo
Source:OpenRouter public model & pricing listingT2observed 11 h agomedium
Capabilities
Modalities
- Modalities
- text
- Input
- text
- Output
- text
Capabilities
Tool calling
Yes
OpenRouter public model & pricing listing · T2
Structured output
Yes
OpenRouter public model & pricing listing · T2
Reasoning
Yes
OpenRouter public model & pricing listing · T2
Vision
Unavailable
Audio
Unavailable
Fine-tuning available
Unavailable
- Context window
Source:OpenRouter public model & pricing listingT2observed 8 h agomedium
- Max output
Source:OpenRouter public model & pricing listingT2observed 11 h agomedium
- Languages
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 10 h agomedium
Benchmarks14
Compare with another model →Comparable same task and conditions · Partially comparable same task, conditions differ (effort, temperature, judge) · Not comparable different variant or metric
No benchmark results recorded
Providers & Pricing5
All offers in the price terminal →USD per 1M tokens as published by each provider (USD). Rows are append-only: every change is kept in the history below.
Price history
Output price · USD / 1M tokens 3 providers
- Fireworks AI
- Z.ai API
- Together AI
- Fireworks AI$4.4 → $6.612 Sept 2026
- Together AIfirst observed $4.411 Sept 2026
- Z.ai API$4.4 → $2.211 Sept 2026
- Z.ai APIfirst observed $4.411 Sept 2026
- Fireworks AIfirst observed $4.411 Sept 2026
Input price · USD / 1M tokens 3 providers
- Fireworks AI
- Z.ai API
- Together AI
- Fireworks AI$1.4 → $2.112 Sept 2026
- Together AIfirst observed $1.411 Sept 2026
- Z.ai API$1.4 → $0.7011 Sept 2026
- Z.ai APIfirst observed $1.411 Sept 2026
- Fireworks AIfirst observed $1.411 Sept 2026
Hardware fit37
Assumptions (6)
- Estimated, not measured: weights = parameters × bytes/param × 1.15 runtime overhead.
- bytes/param: 4bit = 0.5, 8bit = 1.0, fp16 = 2.0 (uniform quantization, no per-layer exceptions).
- KV cache approximated at 0.5 GB per 8 192 tokens of context, independent of architecture (GQA/MLA models need less).
- A model 'fits' when the estimate is at most the device memory minus 2 GB reserved for the OS and framework.
- Mixture-of-experts models are estimated on total parameters (all experts must be resident); active parameters are ignored.
- Device memory uses the largest configuration when several are listed (e.g. Apple silicon tiers).
Lineage
Open in Graph →Papers1
- arXiv:2602.15763Active35
Timeline24
Full timeline →Fireworks AI lists GLM 5.3 at $2.1 in / $6.6 out per 1M tokens
fireworks_pricingGLM 5.3: context length changed from 1000000 to 1310720
Context window1M tokens→1.31M tokensopenrouterGLM 5.3 scores 83.9% on Terminal-Bench
artificial_analysisGLM 5.3 scores 41.92% on Terminal-Bench
artificial_analysisGLM 5.3 scores 42.26% on Humanity's Last Exam
artificial_analysisGLM 5.3 scores 44.86 on Artificial Analysis Intelligence Index
artificial_analysisGLM 5.3: context length changed from 1310720 to 1000000
Context window1.31M tokens→1M tokensartificial_analysisTogether AI lists GLM 5.3 at $1.4 in / $4.4 out per 1M tokens
together_pricingGLM 5.3: release date changed from 2026-08-25 to 2026-08-18
Release date25 Aug 2026→18 Aug 2026openrouterGLM 5.3: release date changed from 2026-08-18 to 2026-08-25
Release date18 Aug 2026→25 Aug 2026huggingfaceZ.ai API lists GLM 5.3 at $0.7 in / $2.2 out per 1M tokens
openrouterZ.ai API lists GLM 5.3 at $1.4 in / $4.4 out per 1M tokens
openrouterFireworks AI lists GLM 5.3 at $1.4 in / $4.4 out per 1M tokens
fireworks_pricing
Change history46
Viewing AI Atlas as of 12 Sept 2026 — attributes exactly as the atlas knew them on that day; later corrections are not shown.
Back to today →Attributes as of 12 Sept 2026 40 claims in force
- Release date
- 25 Aug 2026
- Openness
- open-weights
- License
- Other
- Architecture
- GlmMoeDsaForCausalLM
- Parameters
- 753.3B
- Context window
- 1.31M tokens
- Max output
- 943.7K tokens
- Modalities
- text
- Input modalities
- text
- Output modalities
- text
- Languages
- en, zh
- Quantization format
- fp8
- File size
- 755.6 GB
- Hugging Face repo
- zai-org/GLM-5.3
- Pipeline tag
- text-generation
- Model card
- Downloads
- 597,626
- Likes
- 1,808
- Aa context window
- 1,000,000
- Aa openness
- open-weights
- Gated
- No
- Hf inference providers
- baseten, deepinfra, fireworks-ai, novita, together, zai-org
- Is quantized
- Yes
- Last modified
- 2026-09-04T06:41:23+00:00
- Library name
- transformers
- License name
- glm-5.3
- Livebench hf link
- https://huggingface.co/zai-org/GLM-5.3
- Aa median output tokens per second
- 62.3
- Downloads all time
- 597,626
- Model type
- glm_moe_dsa
- Openrouter id
- z-ai/glm-5.3
- Openrouter listed at
- 18 Aug 2026
- Reasoning
- Yes
- Structured output
- Yes
- Supported parameters
- frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, parallel_tool_calls, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p
- Tags
- transformers, safetensors, glm_moe_dsa, text-generation, en, zh, fp8
- Tool calling
- Yes
- Weights available
- Yes
- Weights dtype
- BF16, F32, F8_E4M3
Release daterelease_date6
Opennessopenness1
Licenselicense1
Architecturearchitecture1
Parametersparameter_count1
Context windowcontext_length3
Max outputmax_output_tokens1
Modalitiesmodalities1
Input modalitiesmodalities_input1
Output modalitiesmodalities_output1
Languageslanguages1
Quantization formatquant_format1
File sizefile_size_gb1
Hugging Face repohf_repo1
Pipeline tagpipeline_tag1
Model cardmodel_card_url1
Downloadsmetric.downloads1
Likesmetric.likes4
Descriptiondescription1
Gatedgated1
Hf inference providershf_inference_providers1
Is quantizedis_quantized1
Last modifiedlast_modified1
Library namelibrary_name1
License namelicense_name1
Livebench hf linklivebench_hf_link1
Aa median output tokens per secondmetric.aa_median_output_tokens_per_second1
Downloads all timemetric.downloads_all_time1
Model typemodel_type1
Openrouter idopenrouter_id1
Reasoningreasoning1
Structured outputstructured_output1
Supported parameterssupported_parameters1
Tool callingtool_calling1
Weights dtypeweights_dtype1
Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →
Provenance
Attributed facts
37
Source tiers
T237
Freshest observation
4 h ago
Conflicts
None
Source documents 9
Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.
Data quality (72/100) measures how well AI Atlas knows this entity — completeness, primary-source ratio, freshness, conflicts — never how good the model is. Methodology →