Model profile / LG AI Research

Exaone 4.0 32B

Exaone 4.0 32B is a open weight model variant recorded in the BenchLM public dataset.

Release date unavailableOpen weights

Sources & details

Four areas of evidence

The capabilities behind the model.

How the indices work

Reasoning

101.4Index points

Mathematical and scientific problem solving.

This model│ Other models · 145 rated

Score and research range

Full-precision point: 101.39656421591724
Range: 98.78–104.02 · 90% Research Range
Scope: Conditional fuller R0 A

Score includes qualified prediction of missing evidence.

Coding

No current rating

Software engineering and programming.

Awaiting a qualified category score

General

No current rating

Factual reliability and instruction following.

Awaiting a qualified category score

Agentic

No current rating

Tool use, workflows and professional tasks.

Awaiting a qualified category score

13catalogue benchmarks

5catalogue sources

Evidence availableNo score is inferred from missing evidence.

Current index snapshot · 2026-09-13. Each category uses its own scale; scores across categories are not directly comparable. Counts describe published catalogue evidence. A point lead does not establish superiority.

Performance in context

Capabilities, price & speed.

Explore every model

Exaone 4.0 32B has no current Capabilities rating. Explore the rated models below.

Lumina Capabilities Index

Capabilities versus API price

The current Capabilities cohort, with available operational measurements.

Lumina

2025+ · 61 models

Chart loads as you explore

Lumina Capabilities Index

Capabilities versus output speed

The current Capabilities cohort, with available operational measurements.

Lumina

2025+ · 69 models · Developer API · 10K

Chart loads as you explore

Explore the detail

Sources & model details.

Original evidence, with its context.

Current index evidence 1Scores, research ranges and the sources behind each category.

Reasoning

Score includes qualified prediction of missing evidence.

Exact checkpoint: exaone-4-0-32b
Method: methods-a4471cee84fd478d2ed3
Range interpretation: Conditional fuller R0 A
Sources: aa
Source cutoff: {"asOfMeaning":"Cutoff means selected immutable source snapshot, not universal freshness.","newIdentityRecovery":"2026-09-13T08:41:08.804270+00:00","publicProjectionDate":"2026-09-13T09:04:50.275277+00:00","scoring":"Inherited pinned baseline source vintages plus retained4A/4A.1 originals and SWE owner correction; heterogeneous execution dates, often unknown. Not a claim of simultaneous2026-09-13 execution."}

Explore this index

Benchmark measurements 37Published values in each benchmark’s own units.

37 measurements

Exaone 4.0 32B published benchmark measurements
Benchmark	Result	Published configuration	Source
GPQA Diamond2026	62.8%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 21 July 2026 ↗Observed 2026-07-21 · checked 2026-07-21source-checked · reference-only
GPQA Diamond2026	62.8%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-07-27 ↗Observed 2026-07-27 · checked 2026-07-27source-checked · reference-only
GPQA Diamond2026	62.8%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-08-01 ↗Observed 2026-08-01 · checked 2026-08-01source-checked · reference-only
Humanity's Last Exam2026	4.9%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 21 July 2026 ↗Observed 2026-07-21 · checked 2026-07-21source-checked · reference-only
Humanity's Last Exam2026	4.9%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-07-27 ↗Observed 2026-07-27 · checked 2026-07-27source-checked · reference-only
Humanity's Last Exam2026	4.9%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-08-01 ↗Observed 2026-08-01 · checked 2026-08-01source-checked · reference-only
IFBench2026	33.5%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 21 July 2026 ↗Observed 2026-07-21 · checked 2026-07-21source-checked · reference-only
IFBench2026	33.5%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-07-27 ↗Observed 2026-07-27 · checked 2026-07-27source-checked · reference-only
IFBench2026	33.5%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-08-01 ↗Observed 2026-08-01 · checked 2026-08-01source-checked · reference-only
AA-Omniscience2026	-62.3 index	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 21 July 2026 ↗Observed 2026-07-21 · checked 2026-07-21source-checked · reference-only

Showing 10 of 37 measurements

Specifications & source recordsDocumented model details and original references.

Model ID / checkpoint: exaone-4-0-32b
Weight access: Open weights
Context window: 128,000
Maximum output: Not documented
Modalities: text
Details checked: 18 Aug 2026

BenchLM public datasets — 21 July 2026BenchLM · Published 21 Jul 2026 · checked 21 Jul 2026 · source-checked BenchLM public datasets — 2026-07-27BenchLM · Published 27 Jul 2026 · checked 27 Jul 2026 · source-checked Artificial Analysis LLM data API — 2026-07-27Artificial Analysis · Published 27 Jul 2026 · checked 27 Jul 2026 · source-checked BenchLM public datasets — 2026-08-01BenchLM · Published 1 Aug 2026 · checked 1 Aug 2026 · source-checked LiveCodeBench permanent refresh sourceLiveCodeBench · checked 7 Aug 2026 · source-checked

Specialist task evidenceSeparate agent runs and execution conditions.

No specialist task observations retained for this model.

All models

Model profile / LG AI Research

Exaone 4.0 32B

Exaone 4.0 32B is a open weight model variant recorded in the BenchLM public dataset.

Release date unavailableOpen weights

Sources & details

Four areas of evidence

The capabilities behind the model.

How the indices work

Reasoning

101.4Index points

Mathematical and scientific problem solving.

This model│ Other models · 145 rated

Score and research range

Full-precision point: 101.39656421591724
Range: 98.78–104.02 · 90% Research Range
Scope: Conditional fuller R0 A

Score includes qualified prediction of missing evidence.

Coding

No current rating

Software engineering and programming.

Awaiting a qualified category score

General

No current rating

Factual reliability and instruction following.

Awaiting a qualified category score

Agentic

No current rating

Tool use, workflows and professional tasks.

Awaiting a qualified category score

13catalogue benchmarks

5catalogue sources

Evidence availableNo score is inferred from missing evidence.

Performance in context

Capabilities, price & speed.

Explore every model

Exaone 4.0 32B has no current Capabilities rating. Explore the rated models below.

Lumina Capabilities Index

Capabilities versus API price

The current Capabilities cohort, with available operational measurements.

Lumina

2025+ · 61 models

Chart loads as you explore

Lumina Capabilities Index

Capabilities versus output speed

The current Capabilities cohort, with available operational measurements.

Lumina

2025+ · 69 models · Developer API · 10K

Chart loads as you explore

Explore the detail

Sources & model details.

Original evidence, with its context.

Current index evidence 1Scores, research ranges and the sources behind each category.

Reasoning

Score includes qualified prediction of missing evidence.

Exact checkpoint: exaone-4-0-32b
Method: methods-a4471cee84fd478d2ed3
Range interpretation: Conditional fuller R0 A
Sources: aa
Source cutoff: {"asOfMeaning":"Cutoff means selected immutable source snapshot, not universal freshness.","newIdentityRecovery":"2026-09-13T08:41:08.804270+00:00","publicProjectionDate":"2026-09-13T09:04:50.275277+00:00","scoring":"Inherited pinned baseline source vintages plus retained4A/4A.1 originals and SWE owner correction; heterogeneous execution dates, often unknown. Not a claim of simultaneous2026-09-13 execution."}

Explore this index

Benchmark measurements 37Published values in each benchmark’s own units.

37 measurements

Exaone 4.0 32B published benchmark measurements
Benchmark	Result	Published configuration	Source
GPQA Diamond2026	62.8%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 21 July 2026 ↗Observed 2026-07-21 · checked 2026-07-21source-checked · reference-only
GPQA Diamond2026	62.8%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-07-27 ↗Observed 2026-07-27 · checked 2026-07-27source-checked · reference-only
GPQA Diamond2026	62.8%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-08-01 ↗Observed 2026-08-01 · checked 2026-08-01source-checked · reference-only
Humanity's Last Exam2026	4.9%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 21 July 2026 ↗Observed 2026-07-21 · checked 2026-07-21source-checked · reference-only
Humanity's Last Exam2026	4.9%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-07-27 ↗Observed 2026-07-27 · checked 2026-07-27source-checked · reference-only
Humanity's Last Exam2026	4.9%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-08-01 ↗Observed 2026-08-01 · checked 2026-08-01source-checked · reference-only
IFBench2026	33.5%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 21 July 2026 ↗Observed 2026-07-21 · checked 2026-07-21source-checked · reference-only
IFBench2026	33.5%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-07-27 ↗Observed 2026-07-27 · checked 2026-07-27source-checked · reference-only
IFBench2026	33.5%	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 2026-08-01 ↗Observed 2026-08-01 · checked 2026-08-01source-checked · reference-only
AA-Omniscience2026	-62.3 index	Exact BenchLM registry variant Exaone 4.0 32B; bulk export does not retain a complete upstream harness configuration.Source-native system	BenchLM public datasets — 21 July 2026 ↗Observed 2026-07-21 · checked 2026-07-21source-checked · reference-only

Showing 10 of 37 measurements

Specifications & source recordsDocumented model details and original references.

Model ID / checkpoint: exaone-4-0-32b
Weight access: Open weights
Context window: 128,000
Maximum output: Not documented
Modalities: text
Details checked: 18 Aug 2026

Specialist task evidenceSeparate agent runs and execution conditions.

No specialist task observations retained for this model.

The capabilities behind the model.

Capabilities, price & speed.

Capabilities versus API price

Capabilities versus output speed

Sources & model details.

Reasoning

More from LG AI Research.

EXAONE 4.5 33B

K-EXAONE (Non-reasoning)

K-EXAONE (Reasoning)

The capabilities behind the model.

Capabilities, price & speed.

Capabilities versus API price

Capabilities versus output speed

Sources & model details.

Reasoning

More from LG AI Research.

EXAONE 4.5 33B

K-EXAONE (Non-reasoning)

K-EXAONE (Reasoning)