Qwen3 30B A3B Instruct 2507 vs Qwen3.5-Flash
Qwen3 30B A3B Instruct 2507 and Qwen3.5-Flash cost about the same per output token ($0.30 vs $0.26 per million). Updated October 23, 2026; 32 benchmarks have a result for both.
Qwen3 30B A3B Instruct 2507 or Qwen3.5-Flash: which to pick
Only Qwen3.5-Flash has a coding score, 41.1 on the Vibe Coding Index; Qwen3 30B A3B Instruct 2507 has none, so the two cannot be ranked on coding.
Pick Qwen3 30B A3B Instruct 2507 for work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.
- Shared results
- 32
- Qwen3 30B A3B Instruct 2507 only
- 13
- Qwen3.5-Flash only
- 23
- 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
No call
Only Qwen3.5-Flash has a coding score, so no call is made.
One side not measured
Intelligence
AA Intelligence Index
No call
Only Qwen3 30B A3B Instruct 2507 has a AA Intelligence Index result, 7.5.
Directional only
Reasoning
GPQA Diamond
No call
Only Qwen3.5-Flash has a GPQA Diamond result, 82.8%.
Directional only
Math
AIME
No call
Only Qwen3.5-Flash has a AIME result, 92.5%.
Directional only
Knowledge
MMLU Pro
No call
Only Qwen3.5-Flash has a MMLU Pro result, 84.1%.
Directional only
Long documents
Listed context window
Qwen3.5-Flash
Qwen3.5-Flash takes 1M tokens against 262K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen3.5-Flash
Qwen3 30B A3B Instruct 2507 $0.0002, Qwen3.5-Flash $0.0002 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only Qwen3.5-Flash has the number, $0.40.
Directional only
At a glance
- Coding index
- Qwen3 30B A3B Instruct 2507 not benchmarked, Qwen3.5-Flash 41.1 from 2 benchmarks.
- Widest category
- Math, 8.2 points apart on average over 3 shared benchmarks.
- Widest single gap
- GPQA Diamond (Epoch AI run): Qwen3 30B A3B Instruct 2507 55.6% against Qwen3.5-Flash 82.3%, 26.7 points apart on a 0 to 100 scale.
- Output price
- Qwen3 30B A3B Instruct 2507 $0.30 against Qwen3.5-Flash $0.26 per million output tokens, 1.2x apart.
- Context window
- Qwen3 30B A3B Instruct 2507 takes 262K tokens, Qwen3.5-Flash 1M.
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 30B A3B Instruct 2507 | Bars | Qwen3.5-Flash | Ties |
|---|---|---|---|---|
| Coding | 2 | 0 | ||
| Reasoning | 0 | 4 | ||
| Math | 0 | 3 | ||
| Knowledge | 0 | 4 | ||
| General | 2 | 17 | ||
| All shared boards | 4 | 28 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index20
| Benchmark | Qwen3 30B A3B Instruct 2507 | Qwen3.5-Flash | Leader |
|---|---|---|---|
| LMArena WebDev | Not measured | 1244 | Only Qwen3.5-Flash measured |
| LiveCodeBench (Vals AI) | Not measured | 83.3% | Only Qwen3.5-Flash measured |
| SWE-bench Verified (Vals AI) | Not measured | 64.4% | Only Qwen3.5-Flash measured |
| LiveCodeBench (Artificial Analysis run) | 70.7% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| SciCode | 33.0% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| Terminal-Bench Hard (Artificial Analysis run) | 6.1% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| ALE-Bench | Not measured | 265.9 | Only Qwen3.5-Flash measured |
| LMArena Text: coding | 1416 | 1412 | Qwen3 30B A3B Instruct 2507 |
| LMArena Text: software and IT | 1417 | 1416 | Qwen3 30B A3B Instruct 2507 |
| LMArena WebDev: React | Not measured | 1229 | Only Qwen3.5-Flash measured |
| LMArena WebDev: HTML | Not measured | 1291 | Only Qwen3.5-Flash measured |
| Neither model has a result on the other 32 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Qwen3 30B A3B Instruct 2507 | Qwen3.5-Flash | Leader |
|---|---|---|---|
| AA Intelligence Index | 7.5 | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| Epoch Capabilities Index | Not measured | 143.5 | Only Qwen3.5-Flash measured |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Qwen3 30B A3B Instruct 2507 | Qwen3.5-Flash | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 28.1% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| Berkeley Function Calling Leaderboard (BFCL) | 41.4% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| Neither model has a result on the other 40 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index04
| Benchmark | Qwen3 30B A3B Instruct 2507 | Qwen3.5-Flash | Leader |
|---|---|---|---|
| GPQA Diamond | Not measured | 82.8% | Only Qwen3.5-Flash measured |
| MMMU Pro (Vals AI) | Not measured | 81.9% | Only Qwen3.5-Flash measured |
| GPQA Diamond (Artificial Analysis run) | 70.7% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| Humanity's Last Exam (Artificial Analysis run) | 10.3% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| GPQA Diamond (Epoch AI run) | 55.6% | 82.3% | Qwen3.5-Flash |
| Chess puzzles | Not measured | 14.0% | Only Qwen3.5-Flash measured |
| Mystery game puzzles | Not measured | 20.0% | Only Qwen3.5-Flash measured |
| LMArena Text: hard prompts | 1398 | 1403 | Qwen3.5-Flash |
| LMArena Text: expert | 1396 | 1407 | Qwen3.5-Flash |
| LMArena Text: hard prompts in English | 1406 | 1409 | Qwen3.5-Flash |
| Neither model has a result on the other 13 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index03
| Benchmark | Qwen3 30B A3B Instruct 2507 | Qwen3.5-Flash | Leader |
|---|---|---|---|
| AIME | Not measured | 92.5% | Only Qwen3.5-Flash measured |
| AA Math Index | 66.3 | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| AIME 2025 (Artificial Analysis run) | 66.3% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| OTIS Mock AIME | 62.2% | 84.4% | Qwen3.5-Flash |
| FrontierMath, tiers 1 to 3 | Not measured | 18.2% | Only Qwen3.5-Flash measured |
| LMArena Text: math | 1395 | 1407 | Qwen3.5-Flash |
| LMArena Text: mathematical work | 1412 | 1413 | Qwen3.5-Flash |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
| Benchmark | Qwen3 30B A3B Instruct 2507 | Qwen3.5-Flash | Leader |
|---|---|---|---|
| MMLU Pro | Not measured | 84.1% | Only Qwen3.5-Flash measured |
| LegalBench (Vals AI) | Not measured | 84.3% | Only Qwen3.5-Flash measured |
| CaseLaw (Vals AI) | Not measured | 56.0% | Only Qwen3.5-Flash measured |
| CorpFin (Vals AI) | Not measured | 63.6% | Only Qwen3.5-Flash measured |
| TaxEval (Vals AI) | Not measured | 72.2% | Only Qwen3.5-Flash measured |
| MortgageTax (Vals AI) | Not measured | 67.4% | Only Qwen3.5-Flash measured |
| MedCode (Vals AI) | Not measured | 33.0% | Only Qwen3.5-Flash measured |
| MMLU Pro (Artificial Analysis run) | 80.5% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| SimpleQA Verified | Not measured | 20.3% | Only Qwen3.5-Flash measured |
| LMArena Text: science | 1393 | 1415 | Qwen3.5-Flash |
| LMArena Text: medicine and healthcare | 1392 | 1407 | Qwen3.5-Flash |
| LMArena Text: legal and government | 1387 | 1397 | Qwen3.5-Flash |
| LMArena Text: business and finance | 1389 | 1394 | Qwen3.5-Flash |
| Neither model has a result on the other 7 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Qwen3 30B A3B Instruct 2507 | Qwen3.5-Flash | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index217
| Benchmark | Qwen3 30B A3B Instruct 2507 | Qwen3.5-Flash | Leader |
|---|---|---|---|
| SAGE (Vals AI) | Not measured | 42.5% | Only Qwen3.5-Flash measured |
| MedScribe (Vals AI) | Not measured | 70.6% | Only Qwen3.5-Flash measured |
| IFBench | 50.7% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| Long context reasoning (AA-LCR) | 61.3% | Not measured | Only Qwen3 30B A3B Instruct 2507 measured |
| DTBench | 67.2% | 82.9% | Qwen3.5-Flash |
| LMCA | 22.4% | 29.1% | Qwen3.5-Flash |
| LMArena Text | 1384 | 1397 | Qwen3.5-Flash |
| LMArena Text: instruction following | 1363 | 1375 | Qwen3.5-Flash |
| LMArena Text: creative writing | 1317 | 1343 | Qwen3.5-Flash |
| LMArena Text: English | 1391 | 1404 | Qwen3.5-Flash |
| LMArena Text: non-English | 1372 | 1385 | Qwen3.5-Flash |
| LMArena Text: Chinese | 1433 | 1446 | Qwen3.5-Flash |
| LMArena Text: German | 1379 | 1391 | Qwen3.5-Flash |
| LMArena Text: Japanese | 1338 | 1369 | Qwen3.5-Flash |
| LMArena Text: Korean | 1331 | 1344 | Qwen3.5-Flash |
| LMArena Text: French | 1419 | 1413 | Qwen3 30B A3B Instruct 2507 |
| LMArena Text: Spanish | 1404 | 1399 | Qwen3 30B A3B Instruct 2507 |
| LMArena Text: Russian | 1369 | 1380 | Qwen3.5-Flash |
| LMArena Text: Polish | 1353 | 1376 | Qwen3.5-Flash |
| LMArena Text: multi-turn | 1378 | 1393 | Qwen3.5-Flash |
| LMArena Text: longer queries | 1378 | 1392 | Qwen3.5-Flash |
| LMArena Text: writing and language | 1336 | 1369 | Qwen3.5-Flash |
| LMArena Text: entertainment and media | 1334 | 1347 | Qwen3.5-Flash |
| 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 30B A3B Instruct 2507 | Qwen3.5-Flash | 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
Qwen3.5-Flash costs less
Repository review
50K input + 3K output tokens
Qwen3.5-Flash costs less
Agent loop
200K input + 10K output tokens
Qwen3.5-Flash 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 30B A3B Instruct 2507 | Qwen3.5-Flash |
|---|---|---|
| Maker | Qwen | Qwen |
| Weights | Open weights | No public weights listed |
| Released | July 29, 2025 | February 25, 2026 |
| Context window | 262K | 1M |
| Input $/Mtok | $0.09 | $0.065 |
| Output $/Mtok | $0.30 | $0.26 |
| Coding index | Not benchmarked | 41.1 |
| Independent measurements behind it | 0 | 2 |
| Cost per solved task | Not computable | $0.40 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Qwen3 30B A3B Instruct 2507 better than Qwen3.5-Flash for coding?
Only Qwen3.5-Flash has a coding score, 41.1 on the Vibe Coding Index; Qwen3 30B A3B Instruct 2507 has none, so the two cannot be ranked on coding. Pick Qwen3 30B A3B Instruct 2507 for work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.
Which is cheaper, Qwen3 30B A3B Instruct 2507 or Qwen3.5-Flash?
Qwen3 30B A3B Instruct 2507 costs $0.09 per million input tokens and $0.30 per million output tokens; Qwen3.5-Flash costs $0.065 and $0.26. Cost per solved task is known for Qwen3.5-Flash only.
Which has the bigger context window, Qwen3 30B A3B Instruct 2507 or Qwen3.5-Flash?
Qwen3.5-Flash, with 1M tokens against 262K.
How many benchmarks is the Qwen3 30B A3B Instruct 2507 vs Qwen3.5-Flash comparison based on?
32 benchmarks have a result for both models as of October 23, 2026. Qwen3 30B A3B Instruct 2507's coding index rests on 0 independent benchmarks and Qwen3.5-Flash's on 2. Where only one model has a result, the page says so instead of naming a winner.