Qwen3 235B A22B Thinking 2507 vs Hy3
Hy3 costs 6.6x less per output token than Qwen3 235B A22B Thinking 2507 ($0.528 vs $3.50 per million). Updated October 22, 2026; 33 benchmarks have a result for both.
Qwen3 235B A22B Thinking 2507 or Hy3: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Qwen3 235B A22B Thinking 2507 for work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.
- Shared results
- 33
- Qwen3 235B A22B Thinking 2507 only
- 13
- Hy3 only
- 11
- Like-for-like categories
- 6 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
No call
Only Qwen3 235B A22B Thinking 2507 has a AA Intelligence Index result, 12.7.
Directional only
Agentic work
AA Agentic Index
No call
Only Qwen3 235B A22B Thinking 2507 has a AA Agentic Index result, 1.3.
Directional only
Web design
Design Arena: websites
Hy3
Hy3 leads, 1230 against 1052.
Same board, both measured
Long documents
Listed context window
Hy3
Hy3 takes 262K tokens against 128K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Hy3
Qwen3 235B A22B Thinking 2507 $0.0022, Hy3 $0.0004 per turn.
List prices
At a glance
- Coding index
- Qwen3 235B A22B Thinking 2507 not benchmarked, Hy3 not benchmarked.
- Widest category
- Web design, 27.9 points apart on average over 6 shared benchmarks.
- Widest single gap
- Design Arena: UI components: Qwen3 235B A22B Thinking 2507 942 against Hy3 1204, 32.8 points apart on a 0 to 100 scale.
- Output price
- Qwen3 235B A22B Thinking 2507 $3.50 against Hy3 $0.528 per million output tokens, 6.6x apart.
- Context window
- Qwen3 235B A22B Thinking 2507 takes 128K tokens, Hy3 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.
| Category | Qwen3 235B A22B Thinking 2507 | Bars | Hy3 | Ties |
|---|---|---|---|---|
| Coding | 0 | 2 | ||
| Reasoning | 1 | 2 | ||
| Math | 0 | 2 | ||
| Knowledge | 1 | 3 | ||
| Web design | 0 | 6 | ||
| General | 0 | 16 | ||
| All shared boards | 2 | 31 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index02
| Benchmark | Qwen3 235B A22B Thinking 2507 | Hy3 | Leader |
|---|---|---|---|
| LMArena WebDev | Not measured | 1507 | Only Hy3 measured |
| AA Coding Index | 22.1 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| SciCode | Not measured | 48.6% | Only Hy3 measured |
| WeirdML | 41.0% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| ALE-Bench | Not measured | 838.6 | Only Hy3 measured |
| LMArena Text: coding | 1424 | 1466 | Hy3 |
| LMArena Text: software and IT | 1431 | 1463 | Hy3 |
| LMArena WebDev: React | Not measured | 1511 | Only Hy3 measured |
| LMArena WebDev: HTML | Not measured | 1487 | Only Hy3 measured |
| Neither model has a result on the other 34 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Qwen3 235B A22B Thinking 2507 | Hy3 | Leader |
|---|---|---|---|
| AA Intelligence Index | 12.7 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Neither model has a result on the other 1 board here | |||
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Qwen3 235B A22B Thinking 2507 | Hy3 | Leader |
|---|---|---|---|
| AA Agentic Index | 1.3 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Vending-Bench 2 | -$11 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Neither model has a result on the other 40 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index12
| Benchmark | Qwen3 235B A22B Thinking 2507 | Hy3 | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | Not measured | 89.7% | Only Hy3 measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 33.5% | Only Hy3 measured |
| GPQA Diamond (Epoch AI run) | 80.1% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Mystery game puzzles | 9.0% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text: hard prompts | 1416 | 1447 | Hy3 |
| LMArena Text: expert | 1463 | 1461 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: hard prompts in English | 1432 | 1455 | Hy3 |
| Neither model has a result on the other 16 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Qwen3 235B A22B Thinking 2507 | Hy3 | Leader |
|---|---|---|---|
| OTIS Mock AIME | 86.7% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text: math | 1412 | 1473 | Hy3 |
| LMArena Text: mathematical work | 1435 | 1482 | Hy3 |
| Neither model has a result on the other 11 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index13
| Benchmark | Qwen3 235B A22B Thinking 2507 | Hy3 | Leader |
|---|---|---|---|
| SimpleQA Verified | 40.4% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text: science | 1433 | 1457 | Hy3 |
| LMArena Text: medicine and healthcare | 1456 | 1452 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: legal and government | 1415 | 1422 | Hy3 |
| LMArena Text: business and finance | 1413 | 1424 | Hy3 |
| Neither model has a result on the other 15 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index06
| Benchmark | Qwen3 235B A22B Thinking 2507 | Hy3 | Leader |
|---|---|---|---|
| Design Arena: websites | 1052 | 1230 | Hy3 |
| Design Arena: UI components | 942 | 1204 | Hy3 |
| Design Arena: game dev | 956 | 1208 | Hy3 |
| Design Arena: data viz | 950 | 1165 | Hy3 |
| Design Arena: 3D | 997 | 1232 | Hy3 |
| Design Arena: all code categories | 1032 | 1227 | Hy3 |
| Design Arena: SVG | Not measured | 1201 | Only Hy3 measured |
| Design Arena: ASCII art | Not measured | 1152 | Only Hy3 measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index016
| Benchmark | Qwen3 235B A22B Thinking 2507 | Hy3 | Leader |
|---|---|---|---|
| Long context reasoning (AA-LCR) | Not measured | 79.0% | Only Hy3 measured |
| DTBench | 80.3% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| LMCA | 29.3% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Fiction.LiveBench | 75.0% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Creative writing (Lech Mazur) | 82.4% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text | 1415 | 1438 | Hy3 |
| LMArena Text: instruction following | 1386 | 1426 | Hy3 |
| LMArena Text: creative writing | 1384 | 1403 | Hy3 |
| LMArena Text: English | 1429 | 1448 | Hy3 |
| LMArena Text: non-English | 1397 | 1426 | Hy3 |
| LMArena Text: Chinese | 1481 | 1485 | Hy3 |
| LMArena Text: German | 1408 | 1449 | Hy3 |
| LMArena Text: Japanese | 1386 | 1397 | Hy3 |
| LMArena Text: Korean | 1375 | 1396 | Hy3 |
| LMArena Text: French | Not measured | 1456 | Only Hy3 measured |
| LMArena Text: Spanish | 1389 | 1457 | Hy3 |
| LMArena Text: Russian | 1401 | 1432 | Hy3 |
| LMArena Text: Polish | 1390 | 1447 | Hy3 |
| LMArena Text: multi-turn | 1406 | 1436 | Hy3 |
| LMArena Text: longer queries | 1400 | 1442 | Hy3 |
| LMArena Text: writing and language | 1388 | 1414 | Hy3 |
| LMArena Text: entertainment and media | 1370 | 1400 | Hy3 |
| Neither model has a result on the other 26 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Qwen3 235B A22B Thinking 2507 | Hy3 | 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 costs less
Repository review
50K input + 3K output tokens
Hy3 costs less
Agent loop
200K input + 10K output tokens
Hy3 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 235B A22B Thinking 2507 | Tencent Hy3 |
|---|---|---|
| Maker | Qwen | Tencent |
| Weights | Open weights | Open weights |
| Released | July 25, 2025 | July 6, 2026 |
| Context window | 128K | 262K |
| Input $/Mtok | $0.45 | $0.132 |
| Output $/Mtok | $3.50 | $0.528 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 1 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Qwen3 235B A22B Thinking 2507 better than Hy3 for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Qwen3 235B A22B Thinking 2507 for work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.
Which is cheaper, Qwen3 235B A22B Thinking 2507 or Hy3?
Qwen3 235B A22B Thinking 2507 costs $0.45 per million input tokens and $3.50 per million output tokens; Hy3 costs $0.132 and $0.528. Cost per solved task is known for neither of them.
Which has the bigger context window, Qwen3 235B A22B Thinking 2507 or Hy3?
Hy3, with 262K tokens against 128K.
How many benchmarks is the Qwen3 235B A22B Thinking 2507 vs Hy3 comparison based on?
33 benchmarks have a result for both models as of October 22, 2026. Qwen3 235B A22B Thinking 2507's coding index rests on 0 independent benchmarks and Hy3's on 1. Where only one model has a result, the page says so instead of naming a winner.