Agentic
Not comparable- Hy4 preview
- 58.2
- Estimated · #30/152
- K-EXAONE 2.0
- Not ranked
- Basis
- BenchAlign lane · 13 vs 2 public rows
- Reading
- Not comparable
Hy4 preview vs K-EXAONE 2.0
Updated September 10, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
4 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
Share or export
Which one for your work
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
Long documents
Prompts that approach the documented context limit
Hy4 preview
Hy4 preview has the larger documented context window.
Confidence: documented
Show secondary and unsupported calls
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
K-EXAONE 2.0 is not ranked on the public lane for coding, so no winner is named for coding.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
K-EXAONE 2.0 is not ranked on the public lane for agentic, so no winner is named for agentic.
Confidence: limited
Chat turn cost
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Repository review cost
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
24 Hy4 preview only
4 Shared
11 K-EXAONE 2.0 only
- Shared results
- 4
- Hy4 preview only
- 24
- K-EXAONE 2.0 only
- 11
- Like-for-like categories
- 0 / 8
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.
Coding
Not comparable- Hy4 preview
- 59.2
- Estimated · #24/151
- K-EXAONE 2.0
- Not ranked
- Basis
- BenchAlign lane · 8 vs 3 public rows
- Reading
- Not comparable
Reasoning
Not comparable- Hy4 preview
- Not ranked
- K-EXAONE 2.0
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Knowledge
Not comparable- Hy4 preview
- 56.5
- Estimated · #47/183
- K-EXAONE 2.0
- Not ranked
- Basis
- BenchAlign lane · 4 vs 4 public rows
- Reading
- Not comparable
Math
Not comparable- Hy4 preview
- Not ranked
- K-EXAONE 2.0
- Not ranked
- Basis
- Provisional lane · 0 vs 2 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- Hy4 preview
- Not ranked
- K-EXAONE 2.0
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Multimodal
Not comparable- Hy4 preview
- 84.8
- Unranked · 1 rankable row
- K-EXAONE 2.0
- Not ranked
- Basis
- Provisional lane · 1 vs 0 weighted rows
- Reading
- Not comparable
Instruction following
Not comparable- Hy4 preview
- Not ranked
- K-EXAONE 2.0
- Not ranked
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
| Category | Hy4 preview | K-EXAONE 2.0 | Basis | Reading |
|---|---|---|---|---|
| Agentic | 58.2Estimated · #30/152 | Not ranked | Not comparableBenchAlign lane · 13 vs 2 public rows | Not comparable |
| Coding | 59.2Estimated · #24/151 | Not ranked | Not comparableBenchAlign lane · 8 vs 3 public rows | Not comparable |
| Reasoning | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Knowledge | 56.5Estimated · #47/183 | Not ranked | Not comparableBenchAlign lane · 4 vs 4 public rows | Not comparable |
| Math | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 2 weighted rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Multimodal | 84.8Unranked · 1 rankable row | Not ranked | Not comparableProvisional lane · 1 vs 0 weighted rows | Not comparable |
| Instruction following | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign) differ from the provisional-lane categories. Unranked scores sit on the lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
Shape of the matched evidence
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
HLE
Knowledge
What each workload costs
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
Chat turn
1K fresh input + 500 output tokens
- Hy4 preview
- Self-hosted; infrastructure cost varies
- Fits in one request
- K-EXAONE 2.0
- Self-hosted; infrastructure cost varies
- Fits in one request
Hy4 preview has no comparable published API token rate. K-EXAONE 2.0 has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- Hy4 preview
- Self-hosted; infrastructure cost varies
- Fits in one request
- K-EXAONE 2.0
- Self-hosted; infrastructure cost varies
- Fits in one request
Hy4 preview has no comparable published API token rate. K-EXAONE 2.0 has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Hy4 preview
- Self-hosted; infrastructure cost varies
- Fits in one request
- Cached-input rate unavailable
- K-EXAONE 2.0
- Self-hosted; infrastructure cost varies
- Fits in one request
- Cached-input rate unavailable
Hy4 preview has no comparable published API token rate. K-EXAONE 2.0 has no comparable published API token rate.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationHy4 previewK-EXAONE 2.0
Context window
Maximum documented context; output-token limits may be lower.
Hy4 preview
K-EXAONE 2.0
API model ID
Hy4 preview
tencent/Hy4-preview
Tencent Hy4 preview model cardK-EXAONE 2.0
LGAI-EXAONE/K-EXAONE-2.0-750B-A37B
LG AI Research K-EXAONE 2.0 model cardCached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Hy4 preview
No comparable hosted API rate
Tencent Hy4 preview model cardK-EXAONE 2.0
No comparable hosted API rate
LG AI Research K-EXAONE 2.0 model cardDocumented inputs
Hy4 preview
Not sourced
K-EXAONE 2.0
Not sourced
Documented outputs
Hy4 preview
Not sourced
K-EXAONE 2.0
Not sourced
Provider availability
Hy4 preview
Not sourced
K-EXAONE 2.0
Not sourced
Reasoning profile
Hy4 preview
Reasoning
K-EXAONE 2.0
Reasoning
Weight access
Hy4 preview
Open Weight
K-EXAONE 2.0
Open Weight
License
Hy4 preview
Open Weight
K-EXAONE 2.0
Open Weight
Release date
Hy4 preview
2026-08-28
K-EXAONE 2.0
2026-07-31
If you already use one of these models
- Deployment change
- The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
- Quality signal
- At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- Hy4 preview has the larger documented window (1M).
Run the same representative tasks against both endpoints before changing production traffic.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence39 rows
Agentic
Terminal-Bench 2.1
Hy4 preview85.4%SourceK-EXAONE 2.043.8%SourceHy4 preview leads this result
CyberGym
Hy4 preview78.4%SourceK-EXAONE 2.0—Not directly comparable
WideResearch
Hy4 preview83.9%SourceK-EXAONE 2.0—Not directly comparable
DRACO
Hy4 preview77.2%SourceK-EXAONE 2.0—Not directly comparable
MCP Atlas
Hy4 preview83.7%SourceK-EXAONE 2.0—Not directly comparable
Toolathlon-Verified
Hy4 preview74.1%SourceK-EXAONE 2.0—Not directly comparable
APEX-Agents
Hy4 preview37.1%SourceK-EXAONE 2.0—Not directly comparable
skillsBench
Hy4 preview62.9%SourceK-EXAONE 2.0—Not directly comparable
JobBench
Hy4 preview61.7%SourceK-EXAONE 2.0—Not directly comparable
Agents' Last Exam
Hy4 preview22.8%SourceK-EXAONE 2.0—Not directly comparable
AutomationBench
Hy4 preview32.1%SourceK-EXAONE 2.0—Not directly comparable
BankerToolBench
Hy4 preview78.6%SourceK-EXAONE 2.0—Not directly comparable
HLE w/ tools
Hy4 preview55.4%SourceK-EXAONE 2.0—Not directly comparable
Claw-Eval
Hy4 preview—K-EXAONE 2.077.7%SourceNot directly comparable
Coding
Terminal-Bench 2.1
Hy4 preview85.4%SourceK-EXAONE 2.043.8%SourceHy4 preview leads this result
SWE-bench Pro
Hy4 preview65.7%SourceK-EXAONE 2.0—Not directly comparable
SWE Multilingual
Hy4 preview82.9%SourceK-EXAONE 2.0—Not directly comparable
DeepSWE
Hy4 preview64.3%SourceK-EXAONE 2.0—Not directly comparable
NL2Repo
Hy4 preview58.9%SourceK-EXAONE 2.0—Not directly comparable
ProgramBench
Hy4 preview17.5%SourceK-EXAONE 2.0—Not directly comparable
PostTrain Bench
Hy4 preview35.6%SourceK-EXAONE 2.0—Not directly comparable
sweMarathon
Hy4 preview31.9%SourceK-EXAONE 2.0—Not directly comparable
SciCode
Hy4 preview—K-EXAONE 2.037.4%SourceNot directly comparable
SWE-bench Verified
Hy4 preview—K-EXAONE 2.068.2%SourceNot directly comparable
Reasoning
CritPt
Hy4 preview16.9%SourceK-EXAONE 2.0—Not directly comparable
Knowledge
GPQA
Hy4 preview92.3%SourceK-EXAONE 2.0—Not directly comparable
GPQA-D
Hy4 preview92.3%SourceK-EXAONE 2.082.2%SourceHy4 preview leads this result
HLE
Hy4 preview55.4%SourceK-EXAONE 2.018.3%SourceHy4 preview leads this result
HLE w/o tools
Hy4 preview43.4%SourceK-EXAONE 2.0—Not directly comparable
MMLU-Pro
Hy4 preview—K-EXAONE 2.083.5%SourceNot directly comparable
MMMLU
Hy4 preview—K-EXAONE 2.086.6%SourceNot directly comparable
Math
Multilingual
PolyMath
Hy4 preview—K-EXAONE 2.071.3%SourceNot directly comparable
Multimodal
OfficeQA Pro
Hy4 preview66.2%SourceK-EXAONE 2.0—Not directly comparable
Instruction following
Frequently asked questions
Which is better, Hy4 preview or K-EXAONE 2.0?
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner. The page therefore keeps the decision tied to the specific documented workload.
Which is better for coding, Hy4 preview or K-EXAONE 2.0?
K-EXAONE 2.0 is not ranked on the public lane for coding, so no winner is named for coding.
Which is better for agentic tasks, Hy4 preview or K-EXAONE 2.0?
K-EXAONE 2.0 is not ranked on the public lane for agentic tasks, so no winner is named for agentic tasks.
Which costs less, Hy4 preview or K-EXAONE 2.0?
Both models do not have comparable published API token rates, so this page does not name a universal price winner.
Which has the larger context window, Hy4 preview or K-EXAONE 2.0?
Hy4 preview has the larger documented context window: 1M, compared with 262K.
Related comparisons
Last updated September 10, 2026
Watch Hy4 preview vs K-EXAONE 2.0
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.