Skip to content
AI Atlas
ModelActiveOpen weights

exaone-4-5-33b

LG AI Researchfamily · EXAONE 4.5huggingface.co/LGAI-EXAONE/EXAONE-4.5-33B

Open in Graph
data quality66

Updated 4 h ago · first seen 11 Sept 2026

model_01M29A5BYYQKHJQWGEFAE46F0J

Overview

Identity

Canonical model
Yesidentity confidence: highOne row per real model release. Artifacts (checkpoints, quantisations, conversions) and folded evaluation variants point here.
Official checkpoints
official_checkpoints = hf_repo identifiers carried by the model itself; artifacts are separate entities pointing here through canonical_id.
Artifacts
3 quantizations0 official · 3 third-partySeparate entities (checkpoint · quantization · conversion · packaging) pointing to this model through canonical_id.
Provider deployments
None recorded
API aliases
exaone-4-5-33bexaone-4-5-33b-non-reasoningIdentifiers under which providers and evaluators refer to this model.
Folded evaluation variants
1Effort / thinking variants (…-high, …-non-reasoning) are result configurations of this model, not separate models. Their old URLs redirect here.

Openness

Open weights— weights downloadable under Other; 7 dimensions unknown.

Weights downloadable under a permissive or Creative Commons licence allowing commercial use; code or data may be missing.

  • Weights

    Yes

  • Inference code

    —

  • Training code

    —

  • Training data

    —

  • Dataset

    —

  • Commercial use

    —

  • Redistribution

    —

  • Derivatives

    —

Licence: Other (unclassified licence) (unknown)

dimensions marked null are unknown, not false

Key facts

Release date

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 d agomedium

Architecture

Architecture

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 d agomedium

Model type

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 d agomedium

Parameters

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 d agomedium

Weights dtype

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 d agomedium

File size

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 d agomedium

Library name

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 d agomedium

Pipeline tag

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 d agomedium

Hugging Face repo

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 d agomedium

Capabilities

Modalities

Modalities
imagetext
Input
imagetext
Output
text

Capabilities

  • Tool calling

    Unavailable

  • Structured output

    Unavailable

  • Reasoning

    Yes

    Artificial Analysis · T2

  • Vision

    Unavailable

  • Audio

    Unavailable

  • Fine-tuning available

    Unavailable

Context window

Source:Artificial AnalysisT2observed 5 d agomedium

Languages

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 d agomedium

Benchmarks16

Compare with another model →

Comparable same task and conditions · Partially comparable same task, conditions differ (effort, temperature, judge) · Not comparable different variant or metric

Benchmark results grouped by comparability group
Benchmark · groupBest scoreTrustConfigurationResultsvs leaderEvaluatedSource
Terminal-Benchagentic · accuracy · variant=hard · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysisvarianthardreasoningonconditions differ across rows → partially comparable2−45.5 ptvs gpt-5.6-solobs. 16 Sept 2026artificialanalysis.aiT2
Terminal-Benchagentic · accuracy · variant=v2.1 · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysisvariantv2.1reasoningonconditions differ across rows → partially comparable2non-primary groupobs. 15 Sept 2026artificialanalysis.aiT2
τ²-benchagentic · pass^1 · variant=Telecom · evaluator=Artificial AnalysisIndependentevaluatorArtificial AnalysisvariantTelecomreasoningonconditions differ across rows → partially comparable1non-primary groupobs. 16 Sept 2026artificialanalysis.aiT2
τ²-benchagentic · pass^1 · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysis1−21.1 ptvs Z.ai GLM 5.2obs. 11 Sept 2026artificialanalysis.aiT2
Artificial Analysis Intelligence Indexcomposite · indexIndependentreasoningonversion4.3conditions differ across rows → partially comparable2−40.2vs Claude Fable 5.1obs. 16 Sept 2026artificialanalysis.aiT2
IFBenchinstruction-following · accuracy · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysisreasoningonconditions differ across rows → partially comparable2−25.4 ptvs Grok 4.3obs. 16 Sept 2026artificialanalysis.aiT2
Humanity's Last Examknowledge · accuracy · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysisreasoningonconditions differ across rows → partially comparable2−46.3 ptvs Claude Fable 5.1obs. 16 Sept 2026artificialanalysis.aiT2
MMMU-Promultimodal · accuracy · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysisreasoningonconditions differ across rows → partially comparable2−19.5 ptvs gpt-6-astraobs. 16 Sept 2026artificialanalysis.aiT2
GPQA Diamondreasoning · accuracy · variant=Diamond · evaluator=Artificial AnalysisIndependentevaluatorArtificial AnalysisvariantDiamondreasoningonconditions differ across rows → partially comparable1non-primary groupobs. 16 Sept 2026artificialanalysis.aiT2
GPQA Diamondreasoning · accuracy · variant=GPQA Diamond · evaluator=Artificial AnalysisIndependentevaluatorArtificial AnalysisvariantGPQA Diamond1−16.9 ptvs gpt-6-astraobs. 11 Sept 2026artificialanalysis.aiT2

Current rows only, grouped by benchmark → canonical metric → comparability group (task configuration). Effort variants folded into this model appear as rows of the same group. 16 current rows in total. “vs leader” compares with the current leader of the benchmark's primary group only; other groups are not directly comparable. Comparability rules →

Hardware fit37

Estimated

36 of 37 device × quantization combinations fit.

Run locally: your machine →
Estimated hardware fit
HardwareQuantizationDevice memoryEst. memoryFits
Apple M3 Ultra4bit—20.3 GB est.Yes
Mac Studio (Apple M5 Ultra)4bit—20.3 GB est.Yes
Apple M2 Ultra4bit—20.3 GB est.Yes
Apple M1 Ultra4bit—20.3 GB est.Yes
Apple M3 Max4bit—20.3 GB est.Yes
Apple M4 Max4bit—20.3 GB est.Yes
Mac Studio (Apple M5 Max)4bit—20.3 GB est.Yes
MacBook Pro (Apple M5 Max)4bit—20.3 GB est.Yes
Apple M2 Max4bit—20.3 GB est.Yes
Apple M1 Max4bit—20.3 GB est.Yes
Apple M4 Pro4bit—20.3 GB est.Yes
Mac mini (Apple M5 Pro)4bit—20.3 GB est.Yes
MacBook Pro (Apple M5 Pro)4bit—20.3 GB est.Yes
Apple M3 Pro4bit—20.3 GB est.Yes
Apple M1 Pro4bit—20.3 GB est.Yes
Apple M2 Pro4bit—20.3 GB est.Yes
Apple M44bit—20.3 GB est.Yes
iMac (Apple M4)4bit—20.3 GB est.Yes
Mac mini (Apple M6)4bit—20.3 GB est.Yes
MacBook Air (Apple M5)4bit—20.3 GB est.Yes
MacBook Pro (Apple M5)4bit—20.3 GB est.Yes
Apple M24bit—20.3 GB est.Yes
Apple M34bit—20.3 GB est.Yes
NVIDIA GeForce RTX 30904bit24 GB20.3 GB est.Yes
NVIDIA GeForce RTX 40904bit24 GB20.3 GB est.Yes
NVIDIA GeForce RTX 50904bit32 GB20.3 GB est.Yes
NVIDIA A100 80GB4bit80 GB20.3 GB est.Yes
NVIDIA H100 SXM4bit80 GB20.3 GB est.Yes
NVIDIA H100 NVL4bit94 GB20.3 GB est.Yes
NVIDIA DGX Spark4bit128 GB20.3 GB est.Yes
NVIDIA H2004bit141 GB20.3 GB est.Yes
NVIDIA H200 NVL4bit141 GB20.3 GB est.Yes
NVIDIA B2004bit180 GB20.3 GB est.Yes
AMD Instinct MI300X4bit192 GB20.3 GB est.Yes
AMD Instinct MI325X4bit256 GB20.3 GB est.Yes
NVIDIA DGX B2004bit1,440 GB20.3 GB est.Yes
Apple M14bit—20.3 GB est.No
Assumptions (7)
  • Estimated, not measured: weights = parameters × bytes/param × 1.15 runtime overhead (or the observed artifact file size when one is recorded).
  • bytes/param: 4bit = 0.5, 8bit = 1.0, fp16 = 2.0 (uniform quantization, no per-layer exceptions).
  • KV cache: 2 × layers × kv_heads × head_dim × 2 bytes × context × batch when the architecture is known; otherwise 0.5 GB per 8 192 tokens (× batch), 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).
  • Multi-GPU: device memories are summed; interconnect bandwidth, tensor-parallel replication and pipeline bubbles are not modelled.

Lineage

Open in Graph →
Explicit derived_from / fine_tuned_from / distilled_from relations stated by sources; artifacts collapsed by kind.
DESCENDANTS 0 · ARTIFACTS3 quantizationsartifacts · collapsed3 quantizations — artifacts · collapsedexaone-4-5-33b34.4B params · this modelexaone-4-5-33b — 34.4B params · this model

    Versions & Artifacts3

    Version history

    Context windowfirst observation only

    11 Sept 2026current

    Licensefirst observation only

    12 Sept 2026current

    Opennessfirst observation only

    11 Sept 2026current

    Parametersfirst observation only

    12 Sept 2026current

    Each hop is a claim: click a value for its source, tier and observation time. Nothing is overwritten — a new observation closes the previous claim.

    Artifacts 3

    quantizations 3

    Papers1

    Timeline9

    Full timeline →

    Change history

    Temporal, append-only claims: a new observation closes the previous claim instead of overwriting it. Rewind the record with the as-of picker.
    1 claims · 1 propertiesShow all properties

    Architecturearchitecture1

    Claim history for Architecture
    ValueValid from → toStatusSourceConfidenceExtractor
    Exaone4_5_ForConditionalGeneration → currentcurrentHugging Face Hub (public pages, model cards, papers)T2mediumdeterministic

    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

    32

    Source tiers

    T232

    Freshest observation

    5 h ago

    Conflicts

    None

    Source documents 7

    Source documents
    SourceDocumentTypeTierLast observedSnapshots
    Hugging Face Hub (public pages, model cards, papers)huggingface.co/LGAI-EXAONE/EXAONE-4.5-33B-GGUF model_pageT2· Quality secondary31 min ago25
    Hugging Face Hub (public pages, model cards, papers)huggingface.co/LGAI-EXAONE/EXAONE-4.5-33B-FP8 model_pageT2· Quality secondary31 min ago25
    Hugging Face Hub (public pages, model cards, papers)huggingface.co/LGAI-EXAONE/EXAONE-4.5-33B-AWQ model_pageT2· Quality secondary31 min ago25
    Hugging Face Hub (public pages, model cards, papers)huggingface.co/LGAI-EXAONE/EXAONE-4.5-33B model_pageT2· Quality secondary31 min ago25
    Hugging Face Hub (public pages, model cards, papers)huggingface.co/models?author=LGAI-EXAONE&p=0&sort=downloads listingT2· Quality secondary53 min ago26
    Hugging Face Hub (public pages, model cards, papers)huggingface.co/LGAI-EXAONE/EXAONE-4.5-33B/raw/main/README.md model_cardT2· Quality secondary4 h ago1
    Artificial Analysisartificialanalysis.ai/leaderboards/models leaderboardT2· Quality secondary6 h ago17

    Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.

    Data quality (66/100) measures how well AI Atlas knows this entity — completeness, primary-source ratio, freshness, conflicts — never how good the model is. Methodology →