Gemma 4 31B vs Qwen3 235B A22B Thinking 2507
Gemma 4 31B leads Qwen3 235B A22B Thinking 2507 on the Artificial Analysis Intelligence Index, 14.7 to 12.7, and Gemma 4 31B costs 8.8x less per output token ($0.40 vs $3.50 per million). Updated October 22, 2026; 33 benchmarks have a result for both.
Gemma 4 31B or Qwen3 235B A22B Thinking 2507: which to pick
Only Gemma 4 31B has a coding score, 39.0 on the Vibe Coding Index; Qwen3 235B A22B Thinking 2507 has none, so the two cannot be ranked on coding.
Pick Qwen3 235B A22B Thinking 2507 for work measured by GPQA Diamond (Epoch AI run) or OTIS Mock AIME, where it leads.
- Shared results
- 33
- Gemma 4 31B only
- 39
- Qwen3 235B A22B Thinking 2507 only
- 13
- 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
Only Gemma 4 31B has a coding score, so no call is made.
One side not measured
Intelligence
AA Intelligence Index
Gemma 4 31B
Gemma 4 31B leads, 14.7 against 12.7.
Same board, both measured
Agentic work
AA Agentic Index
Gemma 4 31B
Gemma 4 31B leads, 4.2 against 1.3.
Same board, both measured
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
Gemma 4 31B
Gemma 4 31B takes 262K tokens against 128K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Gemma 4 31B
Gemma 4 31B $0.0003, Qwen3 235B A22B Thinking 2507 $0.0022 per turn.
List prices
At a glance
- Coding index
- Gemma 4 31B 39.0 from 2 benchmarks, Qwen3 235B A22B Thinking 2507 not benchmarked.
- Widest category
- Coding, 11.0 points apart on average over 4 shared benchmarks.
- Widest single gap
- SimpleQA Verified: Gemma 4 31B 10.4% against Qwen3 235B A22B Thinking 2507 40.4%, 30.0 points apart on a 0 to 100 scale.
- Output price
- Gemma 4 31B $0.40 against Qwen3 235B A22B Thinking 2507 $3.50 per million output tokens, 8.8x apart.
- Context window
- Gemma 4 31B takes 262K tokens, Qwen3 235B A22B Thinking 2507 128K.
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 | Gemma 4 31B | Bars | Qwen3 235B A22B Thinking 2507 | Ties |
|---|---|---|---|---|
| Coding | 4 | 0 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 3 | 1 | ||
| Math | 2 | 1 | ||
| Knowledge | 3 | 2 | ||
| General | 14 | 1 | ||
| All shared boards | 28 | 5 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index40
| Benchmark | Gemma 4 31B | Qwen3 235B A22B Thinking 2507 | Leader |
|---|---|---|---|
| LiveBench coding | 49.0% | Not measured | Only Gemma 4 31B measured |
| LMArena WebDev | 1367 | Not measured | Only Gemma 4 31B measured |
| AA Coding Index | 43.4 | 22.1 | Gemma 4 31B |
| SciCode | 45.5% | Not measured | Only Gemma 4 31B measured |
| Terminal-Bench Hard (Artificial Analysis run) | 36.4% | Not measured | Only Gemma 4 31B measured |
| WeirdML | 52.3% | 41.0% | Gemma 4 31B |
| ALE-Bench | 925.5 | Not measured | Only Gemma 4 31B measured |
| LMArena Text: coding | 1459 | 1424 | Gemma 4 31B |
| LMArena Text: software and IT | 1463 | 1431 | Gemma 4 31B |
| LMArena WebDev: React | 1359 | Not measured | Only Gemma 4 31B measured |
| LMArena WebDev: HTML | 1380 | Not measured | Only Gemma 4 31B measured |
| SciCode | 8.9% | Not measured | Only Gemma 4 31B measured |
| Neither model has a result on the other 31 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Gemma 4 31B | Qwen3 235B A22B Thinking 2507 | Leader |
|---|---|---|---|
| AA Intelligence Index | 14.7 | 12.7 | Gemma 4 31B |
| Epoch Capabilities Index | 143.0 | Not measured | Only Gemma 4 31B measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Gemma 4 31B | Qwen3 235B A22B Thinking 2507 | Leader |
|---|---|---|---|
| AA Agentic Index | 4.2 | 1.3 | Gemma 4 31B |
| LiveBench research | 44.4% | Not measured | Only Gemma 4 31B measured |
| LiveBench browser | 4.0% | Not measured | Only Gemma 4 31B measured |
| LiveBench long agentic | 13.3% | Not measured | Only Gemma 4 31B measured |
| τ²-Bench Telecom | 65.5% | Not measured | Only Gemma 4 31B measured |
| Vending-Bench 2 | Not measured | -$11 | Only Qwen3 235B A22B Thinking 2507 measured |
| Neither model has a result on the other 36 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index31
| Benchmark | Gemma 4 31B | Qwen3 235B A22B Thinking 2507 | Leader |
|---|---|---|---|
| LiveBench reasoning | 68.3% | Not measured | Only Gemma 4 31B measured |
| GPQA Diamond (Artificial Analysis run) | 85.7% | Not measured | Only Gemma 4 31B measured |
| Humanity's Last Exam (Artificial Analysis run) | 23.6% | Not measured | Only Gemma 4 31B measured |
| GPQA Diamond (Epoch AI run) | 75.8% | 80.1% | Qwen3 235B A22B Thinking 2507 |
| Mystery game puzzles | Not measured | 9.0% | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text: hard prompts | 1448 | 1416 | Gemma 4 31B |
| LMArena Text: expert | 1464 | 1463 | Gemma 4 31B |
| LMArena Text: hard prompts in English | 1456 | 1432 | Gemma 4 31B |
| LMArena Vision: diagrams | 1283 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: homework | 1310 | Not measured | Only Gemma 4 31B measured |
| Kagi LLM Benchmark | 63.5% | Not measured | Only Gemma 4 31B measured |
| Neither model has a result on the other 12 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index21
| Benchmark | Gemma 4 31B | Qwen3 235B A22B Thinking 2507 | Leader |
|---|---|---|---|
| OTIS Mock AIME | 73.3% | 86.7% | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: math | 1463 | 1412 | Gemma 4 31B |
| LMArena Text: mathematical work | 1465 | 1435 | Gemma 4 31B |
| Neither model has a result on the other 11 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index32
| Benchmark | Gemma 4 31B | Qwen3 235B A22B Thinking 2507 | Leader |
|---|---|---|---|
| CaseLaw (Vals AI) | 52.6% | Not measured | Only Gemma 4 31B measured |
| MortgageTax (Vals AI) | 61.4% | Not measured | Only Gemma 4 31B measured |
| SimpleQA Verified | 10.4% | 40.4% | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: science | 1457 | 1433 | Gemma 4 31B |
| LMArena Text: medicine and healthcare | 1451 | 1456 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: legal and government | 1438 | 1415 | Gemma 4 31B |
| LMArena Text: business and finance | 1430 | 1413 | Gemma 4 31B |
| FACTS Benchmark Suite | 45.3% | Not measured | Only Gemma 4 31B measured |
| Neither model has a result on the other 12 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
| Benchmark | Gemma 4 31B | Qwen3 235B A22B Thinking 2507 | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1052 | Only Qwen3 235B A22B Thinking 2507 measured |
| Design Arena: UI components | Not measured | 942 | Only Qwen3 235B A22B Thinking 2507 measured |
| Design Arena: game dev | Not measured | 956 | Only Qwen3 235B A22B Thinking 2507 measured |
| Design Arena: data viz | Not measured | 950 | Only Qwen3 235B A22B Thinking 2507 measured |
| Design Arena: 3D | Not measured | 997 | Only Qwen3 235B A22B Thinking 2507 measured |
| Design Arena: all code categories | Not measured | 1032 | Only Qwen3 235B A22B Thinking 2507 measured |
| LiveBench design | 58.4% | Not measured | Only Gemma 4 31B measured |
| Neither model has a result on the other 2 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index141
| Benchmark | Gemma 4 31B | Qwen3 235B A22B Thinking 2507 | Leader |
|---|---|---|---|
| LiveBench data analysis | 53.3% | Not measured | Only Gemma 4 31B measured |
| SAGE (Vals AI) | 55.0% | Not measured | Only Gemma 4 31B measured |
| IFBench | 75.6% | Not measured | Only Gemma 4 31B measured |
| Long context reasoning (AA-LCR) | 69.7% | Not measured | Only Gemma 4 31B measured |
| DTBench | 82.7% | 80.3% | Gemma 4 31B |
| LMCA | 39.3% | 29.3% | Gemma 4 31B |
| Fiction.LiveBench | Not measured | 75.0% | Only Qwen3 235B A22B Thinking 2507 measured |
| Creative writing (Lech Mazur) | Not measured | 82.4% | Only Qwen3 235B A22B Thinking 2507 measured |
| Surface Evolver Bench | 30.6% | Not measured | Only Gemma 4 31B measured |
| LMArena Text | 1443 | 1415 | Gemma 4 31B |
| LMArena Text: instruction following | 1433 | 1386 | Gemma 4 31B |
| LMArena Text: creative writing | 1417 | 1384 | Gemma 4 31B |
| LMArena Vision | 1276 | Not measured | Only Gemma 4 31B measured |
| LMArena Text: English | 1451 | 1429 | Gemma 4 31B |
| LMArena Text: non-English | 1430 | 1397 | Gemma 4 31B |
| LMArena Text: Chinese | 1475 | 1481 | Qwen3 235B A22B Thinking 2507 |
| LMArena Text: German | Not measured | 1408 | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text: Japanese | Not measured | 1386 | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text: Korean | Not measured | 1375 | Only Qwen3 235B A22B Thinking 2507 measured |
| LMArena Text: French | 1436 | Not measured | Only Gemma 4 31B measured |
| LMArena Text: Spanish | 1442 | 1389 | Gemma 4 31B |
| LMArena Text: Russian | 1460 | 1401 | Gemma 4 31B |
| LMArena Text: Polish | 1433 | 1390 | Gemma 4 31B |
| LMArena Text: multi-turn | 1452 | 1406 | Gemma 4 31B |
| LMArena Text: longer queries | 1446 | 1400 | Gemma 4 31B |
| LMArena Text: writing and language | 1432 | 1388 | Gemma 4 31B |
| LMArena Text: entertainment and media | 1405 | 1370 | Gemma 4 31B |
| LMArena Vision: English | 1271 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: Chinese | 1326 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: reading text | 1282 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: humor | 1280 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: captioning | 1218 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: recognizing things | 1226 | Not measured | Only Gemma 4 31B measured |
| LMArena Vision: creative writing | 1285 | Not measured | Only Gemma 4 31B measured |
| LMArena Document | 1425 | Not measured | Only Gemma 4 31B measured |
| EQ-Bench 4 | 1120 | Not measured | Only Gemma 4 31B measured |
| Neither model has a result on the other 12 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemma 4 31B | Qwen3 235B A22B Thinking 2507 | 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
Gemma 4 31B costs less
Repository review
50K input + 3K output tokens
Gemma 4 31B costs less
Agent loop
200K input + 10K output tokens
Gemma 4 31B 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 | Google Gemma 4 31B | Qwen3 235B A22B Thinking 2507 |
|---|---|---|
| Maker | Qwen | |
| Weights | Open weights | Open weights |
| Released | April 2, 2026 | July 25, 2025 |
| Context window | 262K | 128K |
| Input $/Mtok | $0.14 | $0.45 |
| Output $/Mtok | $0.40 | $3.50 |
| Coding index | 39.0 | Not benchmarked |
| Independent measurements behind it | 2 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Gemma 4 31B better than Qwen3 235B A22B Thinking 2507 for coding?
Only Gemma 4 31B has a coding score, 39.0 on the Vibe Coding Index; Qwen3 235B A22B Thinking 2507 has none, so the two cannot be ranked on coding. Pick Qwen3 235B A22B Thinking 2507 for work measured by GPQA Diamond (Epoch AI run) or OTIS Mock AIME, where it leads.
Which is cheaper, Gemma 4 31B or Qwen3 235B A22B Thinking 2507?
Gemma 4 31B costs $0.14 per million input tokens and $0.40 per million output tokens; Qwen3 235B A22B Thinking 2507 costs $0.45 and $3.50. Cost per solved task is known for neither of them.
Which has the bigger context window, Gemma 4 31B or Qwen3 235B A22B Thinking 2507?
Gemma 4 31B, with 262K tokens against 128K.
How many benchmarks is the Gemma 4 31B vs Qwen3 235B A22B Thinking 2507 comparison based on?
33 benchmarks have a result for both models as of October 22, 2026. Gemma 4 31B's coding index rests on 2 independent benchmarks and Qwen3 235B A22B Thinking 2507's on 0. Where only one model has a result, the page says so instead of naming a winner.