LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

o3 Mini vs Hy3 preview

Updated October 21, 2026. 31 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace o3 Mini Replace Hy3 preview
o3 MiniOpenAI
Coding
52.1
Overall
12.5
Output $/M
$4.40
1
Hy3 previewTencent
Coding
22.4
Overall
25.3
Output $/M
$0.60
30
vs boards won of 31 shared

Decision reading

Pick o3 Mini for coding: it scores 52.1 on the Vibe Coding Index against 22.4 for Hy3 preview.

Pick Hy3 preview when output tokens cost 7.3x less, or when the context window is larger, 262K against 200K tokens.

Model Ao3 MiniOpenAI52.1/1002 independent benchmarksNo cost per solved taskModel BHy3 previewTencent22.4/1001 independent benchmarkNo cost per solved task
Shared results
31
o3 Mini only
39
Hy3 preview only
5
Like-for-like categories
7 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
o3 Mini 52.1 from 2 benchmarks, Hy3 preview 22.4 from 1 benchmark.
Widest category
Agentic work, 61.4 points apart on average over 1 shared benchmark.
Widest single gap
τ²-Bench Telecom: o3 Mini 31.3% against Hy3 preview 92.7%, 61.4 points apart on a 0 to 100 scale.
Output price
o3 Mini $4.40 against Hy3 preview $0.60 per million output tokens, 7.3x apart.
Context window
o3 Mini takes 200K tokens, Hy3 preview 262K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
Categoryo3 MiniBarsHy3 previewTies
Coding03
Intelligence01
Agentic work01
Reasoning05
Math02
Knowledge04
General114
All shared boards130
CodingLike-for-likeFour of these boards feed the Vibe Coding Index03
Coding benchmarks for o3 Mini and Hy3 preview
Benchmarko3 MiniHy3 previewLeader
Aider polyglot60.4%Not measuredOnly o3 Mini measured
SWE-bench Verified42.4%Not measuredOnly o3 Mini measured
SWE-bench Lite32.3%Not measuredOnly o3 Mini measured
LMArena WebDevNot measured1357Only Hy3 preview measured
AA Coding IndexNot measured58.8Only Hy3 preview measured
LiveCodeBench (Vals AI)71.5%Not measuredOnly o3 Mini measured
LiveCodeBench (Artificial Analysis run)73.4%Not measuredOnly o3 Mini measured
SciCode42.8%Not measuredOnly o3 Mini measured
Terminal-Bench Hard (Artificial Analysis run)6.8%34.1%Hy3 preview
GSO-Bench1.3%Not measuredOnly o3 Mini measured
WeirdML43.7%Not measuredOnly o3 Mini measured
Cybench22.5%Not measuredOnly o3 Mini measured
LMArena Text: coding13781421Hy3 preview
LMArena Text: software and IT13641423Hy3 preview
LMArena WebDev: ReactNot measured1349Only Hy3 preview measured
LMArena WebDev: HTMLNot measured1356Only Hy3 preview measured
Neither model has a result on the other 22 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for o3 Mini and Hy3 preview
Benchmarko3 MiniHy3 previewLeader
AA Intelligence Index12.525.3Hy3 preview
Epoch Capabilities Index140.2Not measuredOnly o3 Mini measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for o3 Mini and Hy3 preview
Benchmarko3 MiniHy3 previewLeader
AA Agentic IndexNot measured24.1Only Hy3 preview measured
τ²-Bench Telecom31.3%92.7%Hy3 preview
Neither model has a result on the other 37 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index05
Reasoning benchmarks for o3 Mini and Hy3 preview
Benchmarko3 MiniHy3 previewLeader
GPQA Diamond75.5%Not measuredOnly o3 Mini measured
GPQA Diamond (Artificial Analysis run)77.3%86.7%Hy3 preview
Humanity's Last Exam (Artificial Analysis run)12.0%27.8%Hy3 preview
GPQA Diamond (Epoch AI run)77.0%Not measuredOnly o3 Mini measured
ARC-AGI34.5%Not measuredOnly o3 Mini measured
ARC-AGI-23.0%Not measuredOnly o3 Mini measured
SimpleBench22.8%Not measuredOnly o3 Mini measured
Chess puzzles15.0%Not measuredOnly o3 Mini measured
Mystery game puzzles7.0%Not measuredOnly o3 Mini measured
LMArena Text: hard prompts13651413Hy3 preview
LMArena Text: expert13641422Hy3 preview
LMArena Text: hard prompts in English13751416Hy3 preview
Neither model has a result on the other 9 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index02
Math benchmarks for o3 Mini and Hy3 preview
Benchmarko3 MiniHy3 previewLeader
AIME86.5%Not measuredOnly o3 Mini measured
MATH 500 (Vals AI)91.8%Not measuredOnly o3 Mini measured
MGSM (Vals AI)91.3%Not measuredOnly o3 Mini measured
OTIS Mock AIME76.9%Not measuredOnly o3 Mini measured
FrontierMath, tiers 1 to 318.6%Not measuredOnly o3 Mini measured
FrontierMath, tier 40.0%Not measuredOnly o3 Mini measured
MATH level 596.5%Not measuredOnly o3 Mini measured
LMArena Text: math13961424Hy3 preview
LMArena Text: mathematical work13991442Hy3 preview
Neither model has a result on the other 4 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
Knowledge benchmarks for o3 Mini and Hy3 preview
Benchmarko3 MiniHy3 previewLeader
MMLU Pro78.7%Not measuredOnly o3 Mini measured
LegalBench (Vals AI)71.5%Not measuredOnly o3 Mini measured
CorpFin (Vals AI)45.3%Not measuredOnly o3 Mini measured
TaxEval (Vals AI)69.4%Not measuredOnly o3 Mini measured
MedQA (Vals AI)94.8%Not measuredOnly o3 Mini measured
MMLU Pro (Artificial Analysis run)80.2%Not measuredOnly o3 Mini measured
SimpleQA Verified15.3%Not measuredOnly o3 Mini measured
LMArena Text: science13391416Hy3 preview
LMArena Text: medicine and healthcare13301405Hy3 preview
LMArena Text: legal and government13281430Hy3 preview
LMArena Text: business and finance13131397Hy3 preview
Neither model has a result on the other 6 boards here
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Web design benchmarks for o3 Mini and Hy3 preview
Benchmarko3 MiniHy3 previewLeader
Neither model has a result on 9 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index114
General benchmarks for o3 Mini and Hy3 preview
Benchmarko3 MiniHy3 previewLeader
IFBench67.1%63.1%o3 Mini
Long context reasoning (AA-LCR)43.0%64.7%Hy3 preview
DTBench68.8%Not measuredOnly o3 Mini measured
LMCA19.0%Not measuredOnly o3 Mini measured
Fiction.LiveBench50.0%Not measuredOnly o3 Mini measured
Creative writing (Lech Mazur)61.7%Not measuredOnly o3 Mini measured
LMArena Text13371401Hy3 preview
LMArena Text: instruction following13371383Hy3 preview
LMArena Text: creative writing12861339Hy3 preview
LMArena Text: English13511409Hy3 preview
LMArena Text: non-English13191390Hy3 preview
LMArena Text: Chinese13791435Hy3 preview
LMArena Text: German1303Not measuredOnly o3 Mini measured
LMArena Text: Japanese1286Not measuredOnly o3 Mini measured
LMArena Text: Korean1313Not measuredOnly o3 Mini measured
LMArena Text: French13341425Hy3 preview
LMArena Text: Spanish13211399Hy3 preview
LMArena Text: Russian13041399Hy3 preview
LMArena Text: Polish1294Not measuredOnly o3 Mini measured
LMArena Text: multi-turn13201411Hy3 preview
LMArena Text: longer queries13431402Hy3 preview
LMArena Text: writing and language13131370Hy3 preview
LMArena Text: entertainment and media12831352Hy3 preview
Neither model has a result on the other 23 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for o3 Mini and Hy3 preview
Benchmarko3 MiniHy3 previewLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

o3 Mini$0.0033Fits in one request
Hy3 preview$0.0005Fits in one request

Hy3 preview costs less

Repository review

50K input + 3K output tokens

o3 Mini$0.068Fits in one request
Hy3 preview$0.011Fits in one request

Hy3 preview costs less

Agent loop

200K input + 10K output tokens

o3 Mini$0.264Exceeds the context window
Hy3 preview$0.042Fits in one request

Hy3 preview costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of o3 Mini and Hy3 preview
SpecificationOpenAI o3 MiniTencent Hy3 preview
MakerOpenAITencent
WeightsNo public weights listedOpen weights
ReleasedJanuary 31, 2025April 22, 2026
Context window200K262K
Input $/Mtok$1.10$0.18
Output $/Mtok$4.40$0.60
Coding index52.122.4
Independent measurements behind it21
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is o3 Mini better than Hy3 preview for coding?

Pick o3 Mini for coding: it scores 52.1 on the Vibe Coding Index against 22.4 for Hy3 preview. Pick Hy3 preview when output tokens cost 7.3x less, or when the context window is larger, 262K against 200K tokens.

Which is cheaper, o3 Mini or Hy3 preview?

o3 Mini costs $1.10 per million input tokens and $4.40 per million output tokens; Hy3 preview costs $0.18 and $0.60. Cost per solved task is known for neither of them.

Which has the bigger context window, o3 Mini or Hy3 preview?

Hy3 preview, with 262K tokens against 200K.

How many benchmarks is the o3 Mini vs Hy3 preview comparison based on?

31 benchmarks have a result for both models as of October 21, 2026. o3 Mini's coding index rests on 2 independent benchmarks and Hy3 preview's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next