o3 vs Hy4 preview
Hy4 preview costs 3.2x less per output token than o3 ($2.501 vs $8.00 per million). Updated October 22, 2026; 3 benchmarks have a result for both.
o3 or Hy4 preview: which to pick
Only Hy4 preview has a coding score, 44.3 on the Vibe Coding Index; o3 has none, so the two cannot be ranked on coding.
Pick o3 for work measured by MedCode (Vals AI), where it leads.
- Shared results
- 3
- o3 only
- 99
- Hy4 preview only
- 21
- Like-for-like categories
- 2 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
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.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Only Hy4 preview has a coding score, so no call is made.
One side not measured
Intelligence
AA Intelligence Index
No call
Only o3 has a AA Intelligence Index result, 20.2.
Directional only
Reasoning
GPQA Diamond
No call
Only o3 has a GPQA Diamond result, 84.1%.
Directional only
Math
AIME
No call
Only o3 has a AIME result, 85.3%.
Directional only
Knowledge
MMLU Pro
No call
Only o3 has a MMLU Pro result, 85.6%.
Directional only
Web design
Design Arena: websites
No call
Only o3 has a Design Arena: websites result, 1036.
Directional only
General
Vals Index
No call
Only Hy4 preview has a Vals Index result, 53.2%.
Directional only
Long documents
Listed context window
Hy4 preview
Hy4 preview takes 1.05M tokens against 200K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Hy4 preview
o3 $0.0060, Hy4 preview $0.0021 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only o3 has the number, $0.571.
Directional only
At a glance
- Coding index
- o3 not benchmarked, Hy4 preview 44.3 from 3 benchmarks.
- Widest category
- General, 6.9 points apart on average over 1 shared benchmark.
- Widest single gap
- MedScribe (Vals AI): o3 76.7% against Hy4 preview 83.6%, 6.9 points apart on a 0 to 100 scale.
- Output price
- o3 $8.00 against Hy4 preview $2.501 per million output tokens, 3.2x apart.
- Context window
- o3 takes 200K tokens, Hy4 preview 1.05M.
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.
| Category | o3 | Bars | Hy4 preview | Ties |
|---|---|---|---|---|
| Knowledge | 1 | 0 | 1 tie | |
| General | 0 | 1 | ||
| All shared boards | 1 | 1 | 1 tie |
CodingDirectional onlySix of these boards feed the Vibe Coding Index22 boards, none shared
| Benchmark | o3 | Hy4 preview | Leader |
|---|---|---|---|
| Aider polyglot | 81.3% | Not measured | Only o3 measured |
| SWE-bench Verified | 58.4% | Not measured | Only o3 measured |
| SWE-bench Multimodal | 36.0% | Not measured | Only o3 measured |
| LMArena WebDev | Not measured | 1631 | Only Hy4 preview measured |
| LiveCodeBench (Vals AI) | 83.9% | Not measured | Only o3 measured |
| Terminal-Bench 4 (Vals AI) | Not measured | 8.1% | Only Hy4 preview measured |
| Vibe Code Bench (Vals AI) | Not measured | 77.5% | Only Hy4 preview measured |
| Code migration (Vals AI) | Not measured | 47.4% | Only Hy4 preview measured |
| ProgramBench (Vals AI) | Not measured | 0.0% | Only Hy4 preview measured |
| IOI (Vals AI) | Not measured | 59.3% | Only Hy4 preview measured |
| Terminal-Bench 2.1 (Vals AI) | Not measured | 55.1% | Only Hy4 preview measured |
| CyberBench (Vals AI) | Not measured | 65.4% | Only Hy4 preview measured |
| LiveCodeBench (Artificial Analysis run) | 80.8% | Not measured | Only o3 measured |
| Terminal-Bench Hard (Artificial Analysis run) | 37.1% | Not measured | Only o3 measured |
| GSO-Bench | 8.8% | Not measured | Only o3 measured |
| WeirdML | 52.4% | Not measured | Only o3 measured |
| ALE-Bench | 933.5 | Not measured | Only o3 measured |
| LMArena Text: coding | 1407 | Not measured | Only o3 measured |
| LMArena Text: software and IT | 1415 | Not measured | Only o3 measured |
| LMArena WebDev: React | Not measured | 1653 | Only Hy4 preview measured |
| LMArena WebDev: HTML | Not measured | 1620 | Only Hy4 preview measured |
| SciCode | 9.2% | Not measured | Only o3 measured |
| Neither model has a result on the other 17 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | o3 | Hy4 preview | Leader |
|---|---|---|---|
| AA Intelligence Index | 20.2 | Not measured | Only o3 measured |
| Epoch Capabilities Index | 146.7 | Not measured | Only o3 measured |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index15 boards, none shared
| Benchmark | o3 | Hy4 preview | Leader |
|---|---|---|---|
| SREBench (Vals AI) | Not measured | 2.3% | Only Hy4 preview measured |
| Finance Agent (Vals AI) | Not measured | 55.1% | Only Hy4 preview measured |
| Tax Agent Bench (Vals AI) | Not measured | 32.6% | Only Hy4 preview measured |
| Legal Research Bench (Vals AI) | Not measured | 45.2% | Only Hy4 preview measured |
| Harvey's Legal Agent Benchmark (Vals AI) | Not measured | 8.8% | Only Hy4 preview measured |
| Excel Modeling Benchmark (Vals AI) | Not measured | 58.1% | Only Hy4 preview measured |
| BioMysteryBench (Vals AI) | Not measured | 69.3% | Only Hy4 preview measured |
| Terminal-Bench-Science (Vals AI) | Not measured | 1.4% | Only Hy4 preview measured |
| τ²-Bench Telecom | 80.7% | Not measured | Only o3 measured |
| METR time horizons | 65.4% | Not measured | Only o3 measured |
| GDPval | 30.8% | Not measured | Only o3 measured |
| DeepResearch Bench | 45.2% | Not measured | Only o3 measured |
| OSWorld | 23.0% | Not measured | Only o3 measured |
| Berkeley Function Calling Leaderboard (BFCL) | 63.0% | Not measured | Only o3 measured |
| BrowseComp-Plus | 65.9% | Not measured | Only o3 measured |
| Neither model has a result on the other 26 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index17 boards, none shared
| Benchmark | o3 | Hy4 preview | Leader |
|---|---|---|---|
| GPQA Diamond | 84.1% | Not measured | Only o3 measured |
| MMMU Pro (Vals AI) | 80.4% | Not measured | Only o3 measured |
| GPQA Diamond (Artificial Analysis run) | 82.7% | Not measured | Only o3 measured |
| Humanity's Last Exam (Artificial Analysis run) | 20.1% | Not measured | Only o3 measured |
| GPQA Diamond (Epoch AI run) | 81.8% | Not measured | Only o3 measured |
| ARC-AGI | 60.8% | Not measured | Only o3 measured |
| ARC-AGI-2 | 6.5% | Not measured | Only o3 measured |
| SimpleBench | 53.1% | Not measured | Only o3 measured |
| Chess puzzles | 37.0% | Not measured | Only o3 measured |
| Mystery game puzzles | 29.0% | Not measured | Only o3 measured |
| VPCT | 52.0% | Not measured | Only o3 measured |
| LMArena Text: hard prompts | 1402 | Not measured | Only o3 measured |
| LMArena Text: expert | 1402 | Not measured | Only o3 measured |
| LMArena Text: hard prompts in English | 1411 | Not measured | Only o3 measured |
| LMArena Vision: diagrams | 1220 | Not measured | Only o3 measured |
| LMArena Vision: homework | 1234 | Not measured | Only o3 measured |
| Kagi LLM Benchmark | 67.6% | Not measured | Only o3 measured |
| Neither model has a result on the other 5 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index11 boards, none shared
| Benchmark | o3 | Hy4 preview | Leader |
|---|---|---|---|
| AIME | 85.3% | Not measured | Only o3 measured |
| MATH 500 (Vals AI) | 94.6% | Not measured | Only o3 measured |
| MGSM (Vals AI) | 91.8% | Not measured | Only o3 measured |
| AA Math Index | 88.3 | Not measured | Only o3 measured |
| AIME 2025 (Artificial Analysis run) | 88.3% | Not measured | Only o3 measured |
| OTIS Mock AIME | 84.4% | Not measured | Only o3 measured |
| FrontierMath, tiers 1 to 3 | 33.3% | Not measured | Only o3 measured |
| ProofBench | Not measured | 75.0% | Only Hy4 preview measured |
| MATH level 5 | 97.8% | Not measured | Only o3 measured |
| LMArena Text: math | 1426 | Not measured | Only o3 measured |
| LMArena Text: mathematical work | 1423 | Not measured | Only o3 measured |
| Neither model has a result on the other 2 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | o3 | Hy4 preview | Leader |
|---|---|---|---|
| MMLU Pro | 85.6% | Not measured | Only o3 measured |
| LegalBench (Vals AI) | 83.8% | 83.8% | Tie |
| CorpFin (Vals AI) | 59.7% | Not measured | Only o3 measured |
| TaxEval (Vals AI) | 74.6% | Not measured | Only o3 measured |
| MortgageTax (Vals AI) | 65.7% | Not measured | Only o3 measured |
| MedQA (Vals AI) | 96.1% | Not measured | Only o3 measured |
| MedCode (Vals AI) | 47.3% | 43.3% | o3 |
| MMLU Pro (Artificial Analysis run) | 85.3% | Not measured | Only o3 measured |
| SimpleQA Verified | 49.4% | Not measured | Only o3 measured |
| GeoBench | 74.0% | Not measured | Only o3 measured |
| LMArena Text: science | 1429 | Not measured | Only o3 measured |
| LMArena Text: medicine and healthcare | 1448 | Not measured | Only o3 measured |
| LMArena Text: legal and government | 1428 | Not measured | Only o3 measured |
| LMArena Text: business and finance | 1393 | Not measured | Only o3 measured |
| FACTS Benchmark Suite | 44.4% | Not measured | Only o3 measured |
| Neither model has a result on the other 3 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index5 boards, none shared
| Benchmark | o3 | Hy4 preview | Leader |
|---|---|---|---|
| Design Arena: websites | 1036 | Not measured | Only o3 measured |
| Design Arena: UI components | 1017 | Not measured | Only o3 measured |
| Design Arena: game dev | 1034 | Not measured | Only o3 measured |
| Design Arena: data viz | 1200 | Not measured | Only o3 measured |
| Design Arena: all code categories | 1020 | Not measured | Only o3 measured |
| Neither model has a result on the other 4 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | o3 | Hy4 preview | Leader |
|---|---|---|---|
| Vals Index | Not measured | 53.2% | Only Hy4 preview measured |
| Public Benefits Bench (Vals AI) | Not measured | 68.6% | Only Hy4 preview measured |
| SAGE (Vals AI) | 41.8% | Not measured | Only o3 measured |
| MedScribe (Vals AI) | 76.7% | 83.6% | Hy4 preview |
| IFBench | 71.4% | Not measured | Only o3 measured |
| Long context reasoning (AA-LCR) | 74.7% | Not measured | Only o3 measured |
| DTBench | 84.8% | Not measured | Only o3 measured |
| LMCA | 39.7% | Not measured | Only o3 measured |
| Fiction.LiveBench | 88.9% | Not measured | Only o3 measured |
| Creative writing (Lech Mazur) | 83.9% | Not measured | Only o3 measured |
| CL-bench | 17.8% | Not measured | Only o3 measured |
| LMArena Text | 1410 | Not measured | Only o3 measured |
| LMArena Text: instruction following | 1368 | Not measured | Only o3 measured |
| LMArena Text: creative writing | 1359 | Not measured | Only o3 measured |
| LMArena Vision | 1214 | Not measured | Only o3 measured |
| LMArena Text: English | 1417 | Not measured | Only o3 measured |
| LMArena Text: non-English | 1401 | Not measured | Only o3 measured |
| LMArena Text: Chinese | 1437 | Not measured | Only o3 measured |
| LMArena Text: German | 1420 | Not measured | Only o3 measured |
| LMArena Text: Japanese | 1403 | Not measured | Only o3 measured |
| LMArena Text: Korean | 1370 | Not measured | Only o3 measured |
| LMArena Text: French | 1430 | Not measured | Only o3 measured |
| LMArena Text: Spanish | 1395 | Not measured | Only o3 measured |
| LMArena Text: Russian | 1406 | Not measured | Only o3 measured |
| LMArena Text: Polish | 1438 | Not measured | Only o3 measured |
| LMArena Text: multi-turn | 1405 | Not measured | Only o3 measured |
| LMArena Text: longer queries | 1372 | Not measured | Only o3 measured |
| LMArena Text: writing and language | 1372 | Not measured | Only o3 measured |
| LMArena Text: entertainment and media | 1373 | Not measured | Only o3 measured |
| LMArena Vision: English | 1221 | Not measured | Only o3 measured |
| LMArena Vision: Chinese | 1237 | Not measured | Only o3 measured |
| LMArena Vision: reading text | 1218 | Not measured | Only o3 measured |
| LMArena Vision: humor | 1222 | Not measured | Only o3 measured |
| LMArena Vision: captioning | 1210 | Not measured | Only o3 measured |
| LMArena Vision: recognizing things | 1234 | Not measured | Only o3 measured |
| LMArena Vision: creative writing | 1201 | Not measured | Only o3 measured |
| Neither model has a result on the other 10 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | o3 | Hy4 preview | Leader |
|---|---|---|---|
| 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
Hy4 preview costs less
Repository review
50K input + 3K output tokens
Hy4 preview costs less
Agent loop
200K input + 10K output tokens
Hy4 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.
| Specification | OpenAI o3 | Tencent Hy4 preview |
|---|---|---|
| Maker | OpenAI | Tencent |
| Weights | No public weights listed | Open weights |
| Released | April 16, 2025 | August 28, 2026 |
| Context window | 200K | 1.05M |
| Input $/Mtok | $2.00 | $0.834 |
| Output $/Mtok | $8.00 | $2.501 |
| Coding index | Not benchmarked | 44.3 |
| Independent measurements behind it | 0 | 3 |
| Cost per solved task | $0.571 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is o3 better than Hy4 preview for coding?
Only Hy4 preview has a coding score, 44.3 on the Vibe Coding Index; o3 has none, so the two cannot be ranked on coding. Pick o3 for work measured by MedCode (Vals AI), where it leads.
Which is cheaper, o3 or Hy4 preview?
o3 costs $2.00 per million input tokens and $8.00 per million output tokens; Hy4 preview costs $0.834 and $2.501. Cost per solved task is known for o3 only.
Which has the bigger context window, o3 or Hy4 preview?
Hy4 preview, with 1.05M tokens against 200K.
How many benchmarks is the o3 vs Hy4 preview comparison based on?
3 benchmarks have a result for both models as of October 22, 2026. o3's coding index rests on 0 independent benchmarks and Hy4 preview's on 3. Where only one model has a result, the page says so instead of naming a winner.