Qwen3.5-27B vs Hy3 preview
Hy3 preview leads Qwen3.5-27B on the Artificial Analysis Intelligence Index, 25.3 to 22.9, and Hy3 preview costs 2.6x less per output token ($0.60 vs $1.56 per million). Updated October 22, 2026; 34 benchmarks have a result for both.
Qwen3.5-27B or Hy3 preview: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Qwen3.5-27B for work measured by Terminal-Bench Hard (Artificial Analysis run) or LMArena WebDev: React, where it leads.
- Shared results
- 34
- Qwen3.5-27B only
- 19
- Hy3 preview only
- 2
- Like-for-like categories
- 7 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
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
Hy3 preview
Hy3 preview leads, 25.3 against 22.9.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Hy3 preview has a AA Agentic Index result, 24.1.
Directional only
Long documents
Listed context window
No call
Both take 262K tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Hy3 preview
Qwen3.5-27B $0.0010, Hy3 preview $0.0005 per turn.
List prices
At a glance
- Coding index
- Qwen3.5-27B not benchmarked, Hy3 preview not benchmarked.
- Widest category
- General, 3.5 points apart on average over 15 shared benchmarks.
- Widest single gap
- Long context reasoning (AA-LCR): Qwen3.5-27B 77.7% against Hy3 preview 64.7%, 13.0 points apart on a 0 to 100 scale.
- Output price
- Qwen3.5-27B $1.56 against Hy3 preview $0.60 per million output tokens, 2.6x apart.
- Context window
- Both take 262K tokens.
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 | Qwen3.5-27B | Bars | Hy3 preview | Ties |
|---|---|---|---|---|
| Coding | 4 | 2 | ||
| Intelligence | 0 | 1 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 3 | 2 | ||
| Math | 1 | 1 | ||
| Knowledge | 3 | 1 | ||
| General | 11 | 4 | ||
| All shared boards | 23 | 11 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index42
| Benchmark | Qwen3.5-27B | Hy3 preview | Leader |
|---|---|---|---|
| LMArena WebDev | 1358 | 1357 | Qwen3.5-27B |
| AA Coding Index | Not measured | 58.8 | Only Hy3 preview measured |
| Terminal-Bench Hard (Artificial Analysis run) | 32.6% | 34.1% | Hy3 preview |
| WeirdML | 39.5% | Not measured | Only Qwen3.5-27B measured |
| ALE-Bench | 349.4 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: coding | 1427 | 1421 | Qwen3.5-27B |
| LMArena Text: software and IT | 1431 | 1423 | Qwen3.5-27B |
| LMArena WebDev: React | 1346 | 1349 | Hy3 preview |
| LMArena WebDev: HTML | 1386 | 1356 | Qwen3.5-27B |
| Neither model has a result on the other 30 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Qwen3.5-27B | Hy3 preview | Leader |
|---|---|---|---|
| AA Intelligence Index | 22.9 | 25.3 | Hy3 preview |
| Neither model has a result on the other 1 board here | |||
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Qwen3.5-27B | Hy3 preview | Leader |
|---|---|---|---|
| AA Agentic Index | Not measured | 24.1 | Only Hy3 preview measured |
| τ²-Bench Telecom | 93.9% | 92.7% | Qwen3.5-27B |
| Vending-Bench 2 | $202 | Not measured | Only Qwen3.5-27B measured |
| Neither model has a result on the other 38 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index32
| Benchmark | Qwen3.5-27B | Hy3 preview | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 85.8% | 86.7% | Hy3 preview |
| Humanity's Last Exam (Artificial Analysis run) | 23.9% | 27.8% | Hy3 preview |
| LMArena Text: hard prompts | 1415 | 1413 | Qwen3.5-27B |
| LMArena Text: expert | 1427 | 1422 | Qwen3.5-27B |
| LMArena Text: hard prompts in English | 1429 | 1416 | Qwen3.5-27B |
| LMArena Vision: diagrams | 1249 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: homework | 1256 | Not measured | Only Qwen3.5-27B measured |
| Neither model has a result on the other 15 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | Qwen3.5-27B | Hy3 preview | Leader |
|---|---|---|---|
| LMArena Text: math | 1429 | 1424 | Qwen3.5-27B |
| LMArena Text: mathematical work | 1429 | 1442 | Hy3 preview |
| Neither model has a result on the other 11 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index31
| Benchmark | Qwen3.5-27B | Hy3 preview | Leader |
|---|---|---|---|
| LMArena Text: science | 1437 | 1416 | Qwen3.5-27B |
| LMArena Text: medicine and healthcare | 1425 | 1405 | Qwen3.5-27B |
| LMArena Text: legal and government | 1405 | 1430 | Hy3 preview |
| LMArena Text: business and finance | 1399 | 1397 | Qwen3.5-27B |
| Neither model has a result on the other 14 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Qwen3.5-27B | Hy3 preview | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index114
| Benchmark | Qwen3.5-27B | Hy3 preview | Leader |
|---|---|---|---|
| IFBench | 75.6% | 63.1% | Qwen3.5-27B |
| Long context reasoning (AA-LCR) | 77.7% | 64.7% | Qwen3.5-27B |
| DTBench | 82.4% | Not measured | Only Qwen3.5-27B measured |
| LMCA | 34.0% | Not measured | Only Qwen3.5-27B measured |
| LMArena Text | 1409 | 1401 | Qwen3.5-27B |
| LMArena Text: instruction following | 1393 | 1383 | Qwen3.5-27B |
| LMArena Text: creative writing | 1362 | 1339 | Qwen3.5-27B |
| LMArena Vision | 1241 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: English | 1425 | 1409 | Qwen3.5-27B |
| LMArena Text: non-English | 1390 | 1390 | Hy3 preview |
| LMArena Text: Chinese | 1478 | 1435 | Qwen3.5-27B |
| LMArena Text: German | 1392 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: Japanese | 1346 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: Korean | 1359 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: French | 1410 | 1425 | Hy3 preview |
| LMArena Text: Spanish | 1407 | 1399 | Qwen3.5-27B |
| LMArena Text: Russian | 1390 | 1399 | Hy3 preview |
| LMArena Text: Polish | 1375 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: multi-turn | 1410 | 1411 | Hy3 preview |
| LMArena Text: longer queries | 1413 | 1402 | Qwen3.5-27B |
| LMArena Text: writing and language | 1380 | 1370 | Qwen3.5-27B |
| LMArena Text: entertainment and media | 1363 | 1352 | Qwen3.5-27B |
| LMArena Vision: English | 1246 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: Chinese | 1301 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: reading text | 1246 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: humor | 1227 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: captioning | 1170 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: recognizing things | 1210 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: creative writing | 1224 | Not measured | Only Qwen3.5-27B measured |
| Neither model has a result on the other 17 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Qwen3.5-27B | Hy3 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
Hy3 preview costs less
Repository review
50K input + 3K output tokens
Hy3 preview costs less
Agent loop
200K input + 10K output tokens
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.
| Specification | Qwen3.5-27B | Tencent Hy3 preview |
|---|---|---|
| Maker | Qwen | Tencent |
| Weights | Open weights | Open weights |
| Released | February 25, 2026 | April 22, 2026 |
| Context window | 262K | 262K |
| Input $/Mtok | $0.195 | $0.18 |
| Output $/Mtok | $1.56 | $0.60 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 1 | 1 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Qwen3.5-27B better than Hy3 preview for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Qwen3.5-27B for work measured by Terminal-Bench Hard (Artificial Analysis run) or LMArena WebDev: React, where it leads.
Which is cheaper, Qwen3.5-27B or Hy3 preview?
Qwen3.5-27B costs $0.195 per million input tokens and $1.56 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, Qwen3.5-27B or Hy3 preview?
Neither: both take 262K tokens.
How many benchmarks is the Qwen3.5-27B vs Hy3 preview comparison based on?
34 benchmarks have a result for both models as of October 22, 2026. Qwen3.5-27B's coding index rests on 1 independent benchmark and Hy3 preview's on 1. Where only one model has a result, the page says so instead of naming a winner.