Claude Opus 4.8 vs Hy4 preview
Claude Opus 4.8 leads Hy4 preview on coding, 65.0 to 44.3 on the Vibe Coding Index, and Hy4 preview costs 10.0x less per output token ($2.501 vs $25.00 per million). Updated October 22, 2026; 18 benchmarks have a result for both.
vs boards won of 18 shared
Claude Opus 4.8 or Hy4 preview: which to pick
Pick Claude Opus 4.8 for coding: it scores 65.0 on the Vibe Coding Index against 44.3 for Hy4 preview.
Pick Hy4 preview for output tokens 10.0x cheaper ($2.501 vs $25.00 per million) and a larger context window (1.05M against 1M tokens).
Model AClaude Opus 4.8Anthropic65.0/1006 independent benchmarks$2.17 per solved taskModel BHy4 previewTencent44.3/1003 independent benchmarksNo cost per solved task
- Shared results
- 18
- Claude Opus 4.8 only
- 112
- Hy4 preview only
- 6
- Like-for-like categories
- 5 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
Claude Opus 4.8
Claude Opus 4.8 scores 65.0 against 44.3.
6 and 3 benchmarks behind the scores
Intelligence
AA Intelligence Index
No call
Only Claude Opus 4.8 has a AA Intelligence Index result, 41.8.
Directional only
Agentic work
AA Agentic Index
No call
Only Claude Opus 4.8 has a AA Agentic Index result, 41.9.
Directional only
Reasoning
GPQA Diamond
No call
Only Claude Opus 4.8 has a GPQA Diamond result, 92.4%.
Directional only
Knowledge
MMLU Pro
No call
Only Claude Opus 4.8 has a MMLU Pro result, 89.6%.
Directional only
Web design
Design Arena: websites
No call
Only Claude Opus 4.8 has a Design Arena: websites result, 1258.
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 1M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Hy4 preview
Claude Opus 4.8 $0.018, Hy4 preview $0.0021 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only Claude Opus 4.8 has the number, $2.17.
Directional only
At a glance
- Coding index
- Claude Opus 4.8 65.0 from 6 benchmarks, Hy4 preview 44.3 from 3 benchmarks.
- Widest category
- Coding, 8.9 points apart on average over 8 shared benchmarks.
- Widest single gap
- Terminal-Bench 4 (Vals AI): Claude Opus 4.8 23.2% against Hy4 preview 8.1%, 20.2 points apart on a 0 to 100 scale.
- Output price
- Claude Opus 4.8 $25.00 against Hy4 preview $2.501 per million output tokens, 10.0x apart.
- Context window
- Claude Opus 4.8 takes 1M 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 | Claude Opus 4.8 | Bars | Hy4 preview | Ties |
|---|---|---|---|---|
| Coding | 4 | 4 | ||
| Agentic work | 2 | 3 | ||
| Math | 0 | 1 | ||
| Knowledge | 1 | 1 | ||
| General | 1 | 1 | ||
| All shared boards | 8 | 10 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index44
| Benchmark | Claude Opus 4.8 | Hy4 preview | Leader |
|---|---|---|---|
| LiveBench coding | 77.3% | Not measured | Only Claude Opus 4.8 measured |
| LMArena WebDev | 1556 | 1631 | Hy4 preview |
| AA Coding Index | 74.3 | Not measured | Only Claude Opus 4.8 measured |
| LiveCodeBench (Vals AI) | 87.8% | Not measured | Only Claude Opus 4.8 measured |
| SWE-bench Verified (Vals AI) | 88.6% | Not measured | Only Claude Opus 4.8 measured |
| Terminal-Bench 4 (Vals AI) | 23.2% | 8.1% | Claude Opus 4.8 |
| Vibe Code Bench (Vals AI) | 82.7% | 77.5% | Claude Opus 4.8 |
| Code migration (Vals AI) | 47.3% | 47.4% | Hy4 preview |
| ProgramBench (Vals AI) | 1.0% | 0.0% | Claude Opus 4.8 |
| IOI (Vals AI) | Not measured | 59.3% | Only Hy4 preview measured |
| Terminal-Bench 2.1 (Vals AI) | 71.9% | 55.1% | Claude Opus 4.8 |
| SkillsBench (Vals AI) | 59.2% | Not measured | Only Claude Opus 4.8 measured |
| CyberBench (Vals AI) | Not measured | 65.4% | Only Hy4 preview measured |
| SciCode | 54.4% | Not measured | Only Claude Opus 4.8 measured |
| Terminal-Bench Hard (Artificial Analysis run) | 58.3% | Not measured | Only Claude Opus 4.8 measured |
| FrontierCode | 46.5% | Not measured | Only Claude Opus 4.8 measured |
| GSO-Bench | 47.1% | Not measured | Only Claude Opus 4.8 measured |
| WeirdML | 82.9% | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: coding | 1490 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: software and IT | 1488 | Not measured | Only Claude Opus 4.8 measured |
| LMArena WebDev: React | 1557 | 1653 | Hy4 preview |
| LMArena WebDev: HTML | 1543 | 1620 | Hy4 preview |
| SciCode | 12.3% | Not measured | Only Claude Opus 4.8 measured |
| Neither model has a result on the other 16 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Claude Opus 4.8 | Hy4 preview | Leader |
|---|---|---|---|
| AA Intelligence Index | 41.8 | Not measured | Only Claude Opus 4.8 measured |
| Epoch Capabilities Index | 159.0 | Not measured | Only Claude Opus 4.8 measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index23
| Benchmark | Claude Opus 4.8 | Hy4 preview | Leader |
|---|---|---|---|
| AA Agentic Index | 41.9 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: agents building web apps | 1233 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: agents building full-stack apps | 1249 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: agents building mobile apps | 1222 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: agents building Android apps | 1233 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: agents building Godot games | 1231 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: agents building games | 1239 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: HTML slides | 1171 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: PowerPoint slides | 1202 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: spreadsheets | 1303 | Not measured | Only Claude Opus 4.8 measured |
| LiveBench research | 81.8% | Not measured | Only Claude Opus 4.8 measured |
| LiveBench browser | 96.0% | Not measured | Only Claude Opus 4.8 measured |
| LiveBench long agentic | 48.3% | Not measured | Only Claude Opus 4.8 measured |
| SREBench (Vals AI) | Not measured | 2.3% | Only Hy4 preview measured |
| Finance Agent (Vals AI) | 53.9% | 55.1% | Hy4 preview |
| Tax Agent Bench (Vals AI) | 30.0% | 32.6% | Hy4 preview |
| Legal Research Bench (Vals AI) | 43.8% | 45.2% | Hy4 preview |
| Harvey's Legal Agent Benchmark (Vals AI) | Not measured | 8.8% | Only Hy4 preview measured |
| Excel Modeling Benchmark (Vals AI) | 69.4% | 58.1% | Claude Opus 4.8 |
| BioMysteryBench (Vals AI) | Not measured | 69.3% | Only Hy4 preview measured |
| Terminal-Bench-Science (Vals AI) | 4.3% | 1.4% | Claude Opus 4.8 |
| Time Horizon Index (Vals AI) | 11.8% | Not measured | Only Claude Opus 4.8 measured |
| τ²-Bench Telecom | 94.4% | Not measured | Only Claude Opus 4.8 measured |
| APEX-Agents | 48.9% | Not measured | Only Claude Opus 4.8 measured |
| Remote Labor Index | 8.3% | Not measured | Only Claude Opus 4.8 measured |
| OSWorld 2.0 | 20.6% | Not measured | Only Claude Opus 4.8 measured |
| DeepResearch Bench | 50.2% | Not measured | Only Claude Opus 4.8 measured |
| PostTrainBench | 33.8% | Not measured | Only Claude Opus 4.8 measured |
| Vending-Bench 2 | $5,787 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Search | 1204 | Not measured | Only Claude Opus 4.8 measured |
| Neither model has a result on the other 11 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index18 boards, none shared
| Benchmark | Claude Opus 4.8 | Hy4 preview | Leader |
|---|---|---|---|
| GPQA Diamond | 92.4% | Not measured | Only Claude Opus 4.8 measured |
| LiveBench reasoning | 89.2% | Not measured | Only Claude Opus 4.8 measured |
| MMMU Pro (Vals AI) | 86.6% | Not measured | Only Claude Opus 4.8 measured |
| GPQA Diamond (Artificial Analysis run) | 92.0% | Not measured | Only Claude Opus 4.8 measured |
| Humanity's Last Exam (Artificial Analysis run) | 48.7% | Not measured | Only Claude Opus 4.8 measured |
| GPQA Diamond (Epoch AI run) | 91.0% | Not measured | Only Claude Opus 4.8 measured |
| ARC-AGI | 92.5% | Not measured | Only Claude Opus 4.8 measured |
| ARC-AGI-2 | 72.1% | Not measured | Only Claude Opus 4.8 measured |
| SimpleBench | 64.8% | Not measured | Only Claude Opus 4.8 measured |
| Chess puzzles | 52.0% | Not measured | Only Claude Opus 4.8 measured |
| Mystery game puzzles | 36.0% | Not measured | Only Claude Opus 4.8 measured |
| EBR-bench | 28.6% | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: hard prompts | 1482 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: expert | 1503 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: hard prompts in English | 1487 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Vision: diagrams | 1315 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Vision: homework | 1337 | Not measured | Only Claude Opus 4.8 measured |
| Kagi LLM Benchmark | 88.8% | Not measured | Only Claude Opus 4.8 measured |
| Neither model has a result on the other 4 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Claude Opus 4.8 | Hy4 preview | Leader |
|---|---|---|---|
| OTIS Mock AIME | 98.3% | Not measured | Only Claude Opus 4.8 measured |
| FrontierMath, tiers 1 to 3 | 80.0% | Not measured | Only Claude Opus 4.8 measured |
| FrontierMath, tier 4 | 56.1% | Not measured | Only Claude Opus 4.8 measured |
| ProofBench | 69.0% | 75.0% | Hy4 preview |
| LMArena Text: math | 1486 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: mathematical work | 1497 | Not measured | Only Claude Opus 4.8 measured |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | Claude Opus 4.8 | Hy4 preview | Leader |
|---|---|---|---|
| MMLU Pro | 89.6% | Not measured | Only Claude Opus 4.8 measured |
| LegalBench (Vals AI) | 83.6% | 83.8% | Hy4 preview |
| CorpFin (Vals AI) | 66.7% | Not measured | Only Claude Opus 4.8 measured |
| TaxEval (Vals AI) | 75.6% | Not measured | Only Claude Opus 4.8 measured |
| MortgageTax (Vals AI) | 69.9% | Not measured | Only Claude Opus 4.8 measured |
| MedCode (Vals AI) | 53.2% | 43.3% | Claude Opus 4.8 |
| SimpleQA Verified | 53.0% | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: science | 1477 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: medicine and healthcare | 1464 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: legal and government | 1477 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: business and finance | 1458 | Not measured | Only Claude Opus 4.8 measured |
| FACTS Benchmark Suite | 53.5% | Not measured | Only Claude Opus 4.8 measured |
| Neither model has a result on the other 6 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index9 boards, none shared
| Benchmark | Claude Opus 4.8 | Hy4 preview | Leader |
|---|---|---|---|
| Design Arena: websites | 1258 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: UI components | 1255 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: game dev | 1264 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: data viz | 1240 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: 3D | 1222 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: all code categories | 1255 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: SVG | 1195 | Not measured | Only Claude Opus 4.8 measured |
| Design Arena: ASCII art | 1277 | Not measured | Only Claude Opus 4.8 measured |
| LiveBench design | 78.8% | Not measured | Only Claude Opus 4.8 measured |
GeneralLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | Claude Opus 4.8 | Hy4 preview | Leader |
|---|---|---|---|
| Vals Index | Not measured | 53.2% | Only Hy4 preview measured |
| LiveBench data analysis | 66.0% | Not measured | Only Claude Opus 4.8 measured |
| Public Benefits Bench (Vals AI) | 68.1% | 68.6% | Hy4 preview |
| SAGE (Vals AI) | 54.8% | Not measured | Only Claude Opus 4.8 measured |
| MedScribe (Vals AI) | 85.8% | 83.6% | Claude Opus 4.8 |
| Vals Multimodal Index | 70.9% | Not measured | Only Claude Opus 4.8 measured |
| IFBench | 62.2% | Not measured | Only Claude Opus 4.8 measured |
| Long context reasoning (AA-LCR) | 77.7% | Not measured | Only Claude Opus 4.8 measured |
| DTBench | 94.9% | Not measured | Only Claude Opus 4.8 measured |
| LMCA | 57.5% | Not measured | Only Claude Opus 4.8 measured |
| Surface Evolver Bench | 87.5% | Not measured | Only Claude Opus 4.8 measured |
| Furniture assembly | 42.5% | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text | 1461 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: instruction following | 1476 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: creative writing | 1454 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Vision | 1294 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: English | 1468 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: non-English | 1450 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: Chinese | 1506 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: German | 1472 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: Japanese | 1442 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: Korean | 1431 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: French | 1481 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: Spanish | 1465 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: Russian | 1474 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: Polish | 1459 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: multi-turn | 1476 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: longer queries | 1483 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: writing and language | 1461 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Text: entertainment and media | 1440 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Vision: English | 1293 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Vision: Chinese | 1346 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Vision: reading text | 1303 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Vision: humor | 1282 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Vision: creative writing | 1301 | Not measured | Only Claude Opus 4.8 measured |
| LMArena Document | 1475 | Not measured | Only Claude Opus 4.8 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 | Claude Opus 4.8 | 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
Claude Opus 4.8$0.018Fits in one request
Hy4 preview$0.0021Fits in one request
Hy4 preview costs less
Repository review
50K input + 3K output tokens
Claude Opus 4.8$0.325Fits in one request
Hy4 preview$0.049Fits in one request
Hy4 preview costs less
Agent loop
200K input + 10K output tokens
Claude Opus 4.8$1.25Fits in one request
Hy4 preview$0.192Fits in one request
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 | Anthropic Claude Opus 4.8 | Tencent Hy4 preview |
|---|---|---|
| Maker | Anthropic | Tencent |
| Weights | No public weights listed | Open weights |
| Released | May 27, 2026 | August 28, 2026 |
| Context window | 1M | 1.05M |
| Input $/Mtok | $5.00 | $0.834 |
| Output $/Mtok | $25.00 | $2.501 |
| Coding index | 65.0 | 44.3 |
| Independent measurements behind it | 6 | 3 |
| Cost per solved task | $2.17 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Claude Opus 4.8 better than Hy4 preview for coding?
Pick Claude Opus 4.8 for coding: it scores 65.0 on the Vibe Coding Index against 44.3 for Hy4 preview. Pick Hy4 preview for output tokens 10.0x cheaper ($2.501 vs $25.00 per million) and a larger context window (1.05M against 1M tokens).
Which is cheaper, Claude Opus 4.8 or Hy4 preview?
Claude Opus 4.8 costs $5.00 per million input tokens and $25.00 per million output tokens; Hy4 preview costs $0.834 and $2.501. Cost per solved task is known for Claude Opus 4.8 only.
Which has the bigger context window, Claude Opus 4.8 or Hy4 preview?
Hy4 preview, with 1.05M tokens against 1M.
How many benchmarks is the Claude Opus 4.8 vs Hy4 preview comparison based on?
18 benchmarks have a result for both models as of October 22, 2026. Claude Opus 4.8's coding index rests on 6 independent benchmarks and Hy4 preview's on 3. Where only one model has a result, the page says so instead of naming a winner.