Qwen3 235B A22B Thinking 2507 vs Qwen3 Next 80B A3B Instruct
Qwen3 235B A22B Thinking 2507 leads Qwen3 Next 80B A3B Instruct on the Artificial Analysis Intelligence Index, 12.7 to 9.6, and Qwen3 Next 80B A3B Instruct costs 3.2x less per output token ($1.10 vs $3.50 per million). Updated October 23, 2026; 29 benchmarks have a result for both.
Qwen3 235B A22B Thinking 2507 or Qwen3 Next 80B A3B Instruct: 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
- 29
- Qwen3 235B A22B Thinking 2507 only
- 17
- Qwen3 Next 80B A3B Instruct only
- 12
- 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
Qwen3 235B A22B Thinking 2507
Qwen3 235B A22B Thinking 2507 leads, 12.7 against 9.6.
Same board, both measured
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
No call
Only Qwen3 235B A22B Thinking 2507 has a Design Arena: websites result, 1052.
Directional only
Long documents
Listed context window
Qwen3 Next 80B A3B Instruct
Qwen3 Next 80B A3B Instruct takes 262K tokens against 128K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen3 Next 80B A3B Instruct
Qwen3 235B A22B Thinking 2507 $0.0022, Qwen3 Next 80B A3B Instruct $0.0006 per turn.
List prices
At a glance
- Coding index
- Qwen3 235B A22B Thinking 2507 not benchmarked, Qwen3 Next 80B A3B Instruct not benchmarked.
- Widest category
- General, 3.6 points apart on average over 17 shared benchmarks.
- Widest single gap
- Fiction.LiveBench: Qwen3 235B A22B Thinking 2507 75.0% against Qwen3 Next 80B A3B Instruct 55.6%, 19.4 points apart on a 0 to 100 scale.
- Output price
- Qwen3 235B A22B Thinking 2507 $3.50 against Qwen3 Next 80B A3B Instruct $1.10 per million output tokens, 3.2x apart.
- Context window
- Qwen3 235B A22B Thinking 2507 takes 128K tokens, Qwen3 Next 80B A3B Instruct 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 | Qwen3 Next 80B A3B Instruct | Ties |
|---|---|---|---|---|
| Coding | 0 | 2 | ||
| Intelligence | 1 | 0 | ||
| Reasoning | 2 | 1 | ||
| Math | 1 | 1 | ||
| Knowledge | 1 | 3 | ||
| General | 7 | 10 | ||
| All shared boards | 12 | 17 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index02
| Benchmark | Qwen3 235B A22B Thinking 2507 | Qwen3 Next 80B A3B Instruct | Leader |
|---|---|---|---|
| AA Coding Index | 22.1 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| LiveCodeBench (Artificial Analysis run) | Not measured | 78.4% | Only Qwen3 Next 80B A3B Instruct measured |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 9.8% | Only Qwen3 Next 80B A3B Instruct measured |
| WeirdML | 41.0% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text: coding | 1424 | 1441 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: software and IT | 1431 | 1447 | Qwen3 Next 80B A3B Instruct |
| Neither model has a result on the other 37 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Qwen3 235B A22B Thinking 2507 | Qwen3 Next 80B A3B Instruct | Leader |
|---|---|---|---|
| AA Intelligence Index | 12.7 | 9.6 | Qwen3 235B A22B Thinking 2507 |
| Neither model has a result on the other 1 board here | |||
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index3 boards, none shared
| Benchmark | Qwen3 235B A22B Thinking 2507 | Qwen3 Next 80B A3B Instruct | Leader |
|---|---|---|---|
| AA Agentic Index | 1.3 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| τ²-Bench Telecom | Not measured | 41.5% | Only Qwen3 Next 80B A3B Instruct measured |
| Vending-Bench 2 | -$11 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Neither model has a result on the other 39 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index21
| Benchmark | Qwen3 235B A22B Thinking 2507 | Qwen3 Next 80B A3B Instruct | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | Not measured | 75.9% | Only Qwen3 Next 80B A3B Instruct measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 12.6% | Only Qwen3 Next 80B A3B Instruct 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 | 1428 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: expert | 1463 | 1417 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: hard prompts in English | 1432 | 1431 | Qwen3 235B A22B Thinking 2507 |
| Kagi LLM Benchmark | Not measured | 66.7% | Only Qwen3 Next 80B A3B Instruct 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 235B A22B Thinking 2507 | Qwen3 Next 80B A3B Instruct | Leader |
|---|---|---|---|
| AA Math Index | Not measured | 84.3 | Only Qwen3 Next 80B A3B Instruct measured |
| AIME 2025 (Artificial Analysis run) | Not measured | 84.3% | Only Qwen3 Next 80B A3B Instruct measured |
| OTIS Mock AIME | 86.7% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text: math | 1412 | 1440 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: mathematical work | 1435 | 1434 | Qwen3 235B A22B Thinking 2507 |
| Neither model has a result on the other 9 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index13
| Benchmark | Qwen3 235B A22B Thinking 2507 | Qwen3 Next 80B A3B Instruct | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | Not measured | 82.4% | Only Qwen3 Next 80B A3B Instruct measured |
| SimpleQA Verified | 40.4% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text: science | 1433 | 1434 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: medicine and healthcare | 1456 | 1455 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: legal and government | 1415 | 1431 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: business and finance | 1412 | 1422 | Qwen3 Next 80B A3B Instruct |
| Neither model has a result on the other 14 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
| Benchmark | Qwen3 235B A22B Thinking 2507 | Qwen3 Next 80B A3B Instruct | Leader |
|---|---|---|---|
| Design Arena: websites | 1052 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Design Arena: UI components | 942 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Design Arena: game dev | 956 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Design Arena: data viz | 950 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Design Arena: 3D | 998 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Design Arena: all code categories | 1032 | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| Neither model has a result on the other 3 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index710
| Benchmark | Qwen3 235B A22B Thinking 2507 | Qwen3 Next 80B A3B Instruct | Leader |
|---|---|---|---|
| IFBench | Not measured | 60.7% | Only Qwen3 Next 80B A3B Instruct measured |
| Long context reasoning (AA-LCR) | Not measured | 63.7% | Only Qwen3 Next 80B A3B Instruct 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% | 55.6% | Qwen3 235B A22B Thinking 2507 |
| Creative writing (Lech Mazur) | 82.4% | Not measured | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text | 1415 | 1417 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: instruction following | 1386 | 1388 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: creative writing | 1384 | 1333 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: English | 1429 | 1421 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: non-English | 1397 | 1407 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: Chinese | 1481 | 1460 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: German | 1408 | 1417 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: Japanese | 1386 | 1395 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: Korean | 1375 | 1364 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: French | Not measured | 1414 | Only Qwen3 Next 80B A3B Instruct measured |
| LMArena Text: Spanish | 1389 | 1435 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: Russian | 1400 | 1404 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: Polish | 1390 | 1399 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: multi-turn | 1406 | 1416 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: longer queries | 1399 | 1402 | Qwen3 Next 80B A3B Instruct |
| LMArena Text: writing and language | 1388 | 1355 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: entertainment and media | 1370 | 1349 | Qwen3 235B A22B Thinking 2507 |
| Neither model has a result on the other 25 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Qwen3 235B A22B Thinking 2507 | Qwen3 Next 80B A3B Instruct | 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
Qwen3 Next 80B A3B Instruct costs less
Repository review
50K input + 3K output tokens
Qwen3 Next 80B A3B Instruct costs less
Agent loop
200K input + 10K output tokens
Qwen3 Next 80B A3B Instruct 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 | Qwen3 Next 80B A3B Instruct |
|---|---|---|
| Maker | Qwen | Qwen |
| Weights | Open weights | Open weights |
| Released | July 25, 2025 | September 11, 2025 |
| Context window | 128K | 262K |
| Input $/Mtok | $0.45 | $0.10 |
| Output $/Mtok | $3.50 | $1.10 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Qwen3 235B A22B Thinking 2507 better than Qwen3 Next 80B A3B Instruct 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 Qwen3 Next 80B A3B Instruct?
Qwen3 235B A22B Thinking 2507 costs $0.45 per million input tokens and $3.50 per million output tokens; Qwen3 Next 80B A3B Instruct costs $0.10 and $1.10. Cost per solved task is known for neither of them.
Which has the bigger context window, Qwen3 235B A22B Thinking 2507 or Qwen3 Next 80B A3B Instruct?
Qwen3 Next 80B A3B Instruct, with 262K tokens against 128K.
How many benchmarks is the Qwen3 235B A22B Thinking 2507 vs Qwen3 Next 80B A3B Instruct comparison based on?
29 benchmarks have a result for both models as of October 23, 2026. Qwen3 235B A22B Thinking 2507's coding index rests on 0 independent benchmarks and Qwen3 Next 80B A3B Instruct's on 0. Where only one model has a result, the page says so instead of naming a winner.