Qwen-Plus vs Qwen3.5-9B
Qwen3.5-9B costs 5.2x less per output token than Qwen-Plus ($0.15 vs $0.78 per million). Updated October 21, 2026; 4 benchmarks have a result for both.
Qwen-Plus or Qwen3.5-9B: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Qwen-Plus for a larger context window (1M against 262K tokens) and work measured by GPQA Diamond (Epoch AI run) or OTIS Mock AIME, where it leads.
- Shared results
- 4
- Qwen-Plus only
- 24
- Qwen3.5-9B only
- 11
- Like-for-like categories
- 3 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 Qwen3.5-9B has a AA Intelligence Index result, 13.3.
Directional only
Long documents
Listed context window
Qwen-Plus
Qwen-Plus takes 1M tokens against 262K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen3.5-9B
Qwen-Plus $0.0006, Qwen3.5-9B $0.0002 per turn.
List prices
At a glance
- Coding index
- Qwen-Plus not benchmarked, Qwen3.5-9B not benchmarked.
- Widest category
- Math, 43.9 points apart on average over 1 shared benchmark.
- Widest single gap
- OTIS Mock AIME: Qwen-Plus 17.8% against Qwen3.5-9B 61.7%, 43.9 points apart on a 0 to 100 scale.
- Output price
- Qwen-Plus $0.78 against Qwen3.5-9B $0.15 per million output tokens, 5.2x apart.
- Context window
- Qwen-Plus takes 1M tokens, Qwen3.5-9B 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 | Qwen-Plus | Bars | Qwen3.5-9B | Ties |
|---|---|---|---|---|
| Reasoning | 0 | 1 | ||
| Math | 0 | 1 | ||
| General | 1 | 1 | ||
| All shared boards | 1 | 3 |
CodingDirectional onlySix of these boards feed the Vibe Coding Index6 boards, none shared
| Benchmark | Qwen-Plus | Qwen3.5-9B | Leader |
|---|---|---|---|
| AA Coding Index | Not measured | 23.5 | Only Qwen3.5-9B measured |
| SciCode | Not measured | 29.5% | Only Qwen3.5-9B measured |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 24.2% | Only Qwen3.5-9B measured |
| Terminal-Bench | Not measured | 9.2% | Only Qwen3.5-9B measured |
| LMArena Text: coding | 1328 | Not measured | Only Qwen-Plus measured |
| LMArena Text: software and IT | 1343 | Not measured | Only Qwen-Plus measured |
| Neither model has a result on the other 33 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Qwen-Plus | Qwen3.5-9B | Leader |
|---|---|---|---|
| AA Intelligence Index | Not measured | 13.3 | Only Qwen3.5-9B measured |
| Epoch Capabilities Index | Not measured | 139.3 | Only Qwen3.5-9B measured |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Qwen-Plus | Qwen3.5-9B | Leader |
|---|---|---|---|
| τ²-Bench Telecom | Not measured | 86.8% | Only Qwen3.5-9B measured |
| Neither model has a result on the other 40 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Qwen-Plus | Qwen3.5-9B | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | Not measured | 80.6% | Only Qwen3.5-9B measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 14.9% | Only Qwen3.5-9B measured |
| GPQA Diamond (Epoch AI run) | 48.1% | 79.0% | Qwen3.5-9B |
| LMArena Text: hard prompts | 1317 | Not measured | Only Qwen-Plus measured |
| LMArena Text: expert | 1328 | Not measured | Only Qwen-Plus measured |
| LMArena Text: hard prompts in English | 1333 | Not measured | Only Qwen-Plus measured |
| Neither model has a result on the other 16 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Qwen-Plus | Qwen3.5-9B | Leader |
|---|---|---|---|
| OTIS Mock AIME | 17.8% | 61.7% | Qwen3.5-9B |
| MATH level 5 | 65.3% | Not measured | Only Qwen-Plus measured |
| LMArena Text: math | 1326 | Not measured | Only Qwen-Plus measured |
| LMArena Text: mathematical work | 1332 | Not measured | Only Qwen-Plus measured |
| Neither model has a result on the other 9 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index4 boards, none shared
| Benchmark | Qwen-Plus | Qwen3.5-9B | Leader |
|---|---|---|---|
| LMArena Text: science | 1352 | Not measured | Only Qwen-Plus measured |
| LMArena Text: medicine and healthcare | 1350 | Not measured | Only Qwen-Plus measured |
| LMArena Text: legal and government | 1349 | Not measured | Only Qwen-Plus measured |
| LMArena Text: business and finance | 1303 | Not measured | Only Qwen-Plus measured |
| 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 | Qwen-Plus | Qwen3.5-9B | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | Qwen-Plus | Qwen3.5-9B | Leader |
|---|---|---|---|
| IFBench | Not measured | 66.7% | Only Qwen3.5-9B measured |
| Long context reasoning (AA-LCR) | Not measured | 70.0% | Only Qwen3.5-9B measured |
| DTBench | 81.1% | 71.2% | Qwen-Plus |
| LMCA | 24.0% | 24.5% | Qwen3.5-9B |
| LMArena Text | 1326 | Not measured | Only Qwen-Plus measured |
| LMArena Text: instruction following | 1303 | Not measured | Only Qwen-Plus measured |
| LMArena Text: creative writing | 1292 | Not measured | Only Qwen-Plus measured |
| LMArena Text: English | 1339 | Not measured | Only Qwen-Plus measured |
| LMArena Text: non-English | 1310 | Not measured | Only Qwen-Plus measured |
| LMArena Text: Chinese | 1347 | Not measured | Only Qwen-Plus measured |
| LMArena Text: Japanese | 1251 | Not measured | Only Qwen-Plus measured |
| LMArena Text: Russian | 1323 | Not measured | Only Qwen-Plus measured |
| LMArena Text: multi-turn | 1336 | Not measured | Only Qwen-Plus measured |
| LMArena Text: longer queries | 1324 | Not measured | Only Qwen-Plus measured |
| LMArena Text: writing and language | 1296 | Not measured | Only Qwen-Plus measured |
| LMArena Text: entertainment and media | 1270 | Not measured | Only Qwen-Plus measured |
| Neither model has a result on the other 30 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Qwen-Plus | Qwen3.5-9B | 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
Qwen3.5-9B costs less
Repository review
50K input + 3K output tokens
Qwen3.5-9B costs less
Agent loop
200K input + 10K output tokens
Qwen3.5-9B 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 | Qwen-Plus | Qwen3.5-9B |
|---|---|---|
| Maker | Qwen | Qwen |
| Weights | No public weights listed | Open weights |
| Released | February 1, 2025 | March 10, 2026 |
| Context window | 1M | 262K |
| Input $/Mtok | $0.26 | $0.10 |
| Output $/Mtok | $0.78 | $0.15 |
| 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 Qwen-Plus better than Qwen3.5-9B for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Qwen-Plus for a larger context window (1M against 262K tokens) and work measured by GPQA Diamond (Epoch AI run) or OTIS Mock AIME, where it leads.
Which is cheaper, Qwen-Plus or Qwen3.5-9B?
Qwen-Plus costs $0.26 per million input tokens and $0.78 per million output tokens; Qwen3.5-9B costs $0.10 and $0.15. Cost per solved task is known for neither of them.
Which has the bigger context window, Qwen-Plus or Qwen3.5-9B?
Qwen-Plus, with 1M tokens against 262K.
How many benchmarks is the Qwen-Plus vs Qwen3.5-9B comparison based on?
4 benchmarks have a result for both models as of October 21, 2026. Qwen-Plus's coding index rests on 0 independent benchmarks and Qwen3.5-9B's on 0. Where only one model has a result, the page says so instead of naming a winner.