Qwen3.5-27B vs Step 3.7 Flash
Qwen3.5-27B leads Step 3.7 Flash on the Artificial Analysis Intelligence Index, 22.9 to 19.5, and Step 3.7 Flash costs 1.4x less per output token ($1.15 vs $1.56 per million). Updated October 21, 2026; 8 benchmarks have a result for both.
Qwen3.5-27B or Step 3.7 Flash: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Qwen3.5-27B for work measured by Terminal-Bench Hard (Artificial Analysis run) or ALE-Bench, where it leads.
- Shared results
- 8
- Qwen3.5-27B only
- 45
- Step 3.7 Flash only
- 10
- 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
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
Qwen3.5-27B
Qwen3.5-27B leads, 22.9 against 19.5.
Same board, both measured
Web design
Design Arena: websites
No call
Only Step 3.7 Flash has a Design Arena: websites result, 1196.
Directional only
Long documents
Listed context window
No call
Both take 262K tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Step 3.7 Flash
Qwen3.5-27B $0.0010, Step 3.7 Flash $0.0008 per turn.
List prices
At a glance
- Coding index
- Qwen3.5-27B not benchmarked, Step 3.7 Flash not benchmarked.
- Widest category
- Coding, 7.3 points apart on average over 2 shared benchmarks.
- Widest single gap
- ALE-Bench: Qwen3.5-27B 349.4 against Step 3.7 Flash 694.1, 11.5 points apart on a 0 to 100 scale.
- Output price
- Qwen3.5-27B $1.56 against Step 3.7 Flash $1.15 per million output tokens, 1.4x apart.
- Context window
- Both take 262K tokens.
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 | Qwen3.5-27B | Bars | Step 3.7 Flash | Ties |
|---|---|---|---|---|
| Coding | 0 | 2 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 0 | 1 | ||
| Reasoning | 2 | 0 | ||
| General | 2 | 0 | ||
| All shared boards | 5 | 3 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index02
| Benchmark | Qwen3.5-27B | Step 3.7 Flash | Leader |
|---|---|---|---|
| LMArena WebDev | 1358 | Not measured | Only Qwen3.5-27B measured |
| AA Coding Index | Not measured | 39.6 | Only Step 3.7 Flash measured |
| SciCode | Not measured | 43.9% | Only Step 3.7 Flash measured |
| Terminal-Bench Hard (Artificial Analysis run) | 32.6% | 35.6% | Step 3.7 Flash |
| WeirdML | 39.5% | Not measured | Only Qwen3.5-27B measured |
| ALE-Bench | 349.4 | 694.1 | Step 3.7 Flash |
| LMArena Text: coding | 1427 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: software and IT | 1431 | Not measured | Only Qwen3.5-27B measured |
| LMArena WebDev: React | 1346 | Not measured | Only Qwen3.5-27B measured |
| LMArena WebDev: HTML | 1386 | Not measured | Only Qwen3.5-27B measured |
| Neither model has a result on the other 29 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Qwen3.5-27B | Step 3.7 Flash | Leader |
|---|---|---|---|
| AA Intelligence Index | 22.9 | 19.5 | Qwen3.5-27B |
| Neither model has a result on the other 1 board here | |||
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Qwen3.5-27B | Step 3.7 Flash | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 93.9% | 98.5% | Step 3.7 Flash |
| Vending-Bench 2 | $202 | Not measured | Only Qwen3.5-27B measured |
| Neither model has a result on the other 39 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Qwen3.5-27B | Step 3.7 Flash | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 85.8% | 80.9% | Qwen3.5-27B |
| Humanity's Last Exam (Artificial Analysis run) | 23.9% | 21.4% | Qwen3.5-27B |
| LMArena Text: hard prompts | 1415 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: expert | 1427 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: hard prompts in English | 1429 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: diagrams | 1249 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: homework | 1256 | Not measured | Only Qwen3.5-27B measured |
| Neither model has a result on the other 15 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Qwen3.5-27B | Step 3.7 Flash | Leader |
|---|---|---|---|
| LMArena Text: math | 1429 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: mathematical work | 1429 | Not measured | Only Qwen3.5-27B measured |
| Neither model has a result on the other 11 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index4 boards, none shared
| Benchmark | Qwen3.5-27B | Step 3.7 Flash | Leader |
|---|---|---|---|
| LMArena Text: science | 1437 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: medicine and healthcare | 1425 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: legal and government | 1405 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: business and finance | 1399 | Not measured | Only Qwen3.5-27B measured |
| Neither model has a result on the other 14 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | Qwen3.5-27B | Step 3.7 Flash | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1196 | Only Step 3.7 Flash measured |
| Design Arena: UI components | Not measured | 1179 | Only Step 3.7 Flash measured |
| Design Arena: game dev | Not measured | 1158 | Only Step 3.7 Flash measured |
| Design Arena: data viz | Not measured | 1172 | Only Step 3.7 Flash measured |
| Design Arena: 3D | Not measured | 1129 | Only Step 3.7 Flash measured |
| Design Arena: all code categories | Not measured | 1182 | Only Step 3.7 Flash measured |
| Design Arena: SVG | Not measured | 1085 | Only Step 3.7 Flash measured |
| Design Arena: ASCII art | Not measured | 1152 | Only Step 3.7 Flash measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Qwen3.5-27B | Step 3.7 Flash | Leader |
|---|---|---|---|
| IFBench | 75.6% | 67.3% | Qwen3.5-27B |
| Long context reasoning (AA-LCR) | 77.7% | 73.7% | Qwen3.5-27B |
| DTBench | 82.4% | Not measured | Only Qwen3.5-27B measured |
| LMCA | 34.0% | Not measured | Only Qwen3.5-27B measured |
| LMArena Text | 1409 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: instruction following | 1393 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: creative writing | 1362 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision | 1241 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: English | 1425 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: non-English | 1390 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: Chinese | 1478 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: German | 1392 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: Japanese | 1346 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: Korean | 1359 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: French | 1410 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: Spanish | 1407 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: Russian | 1390 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: Polish | 1375 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: multi-turn | 1410 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: longer queries | 1413 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: writing and language | 1380 | Not measured | Only Qwen3.5-27B measured |
| LMArena Text: entertainment and media | 1363 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: English | 1246 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: Chinese | 1301 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: reading text | 1246 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: humor | 1227 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: captioning | 1170 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: recognizing things | 1210 | Not measured | Only Qwen3.5-27B measured |
| LMArena Vision: creative writing | 1224 | Not measured | Only Qwen3.5-27B measured |
| Neither model has a result on the other 17 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Qwen3.5-27B | Step 3.7 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
Step 3.7 Flash costs less
Repository review
50K input + 3K output tokens
Step 3.7 Flash costs less
Agent loop
200K input + 10K output tokens
Step 3.7 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 | Qwen3.5-27B | StepFun Step 3.7 Flash |
|---|---|---|
| Maker | Qwen | StepFun |
| Weights | Open weights | Open weights |
| Released | February 25, 2026 | May 28, 2026 |
| Context window | 262K | 262K |
| Input $/Mtok | $0.195 | $0.20 |
| Output $/Mtok | $1.56 | $1.15 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 1 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Qwen3.5-27B better than Step 3.7 Flash for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Qwen3.5-27B for work measured by Terminal-Bench Hard (Artificial Analysis run) or ALE-Bench, where it leads.
Which is cheaper, Qwen3.5-27B or Step 3.7 Flash?
Qwen3.5-27B costs $0.195 per million input tokens and $1.56 per million output tokens; Step 3.7 Flash costs $0.20 and $1.15. Cost per solved task is known for neither of them.
Which has the bigger context window, Qwen3.5-27B or Step 3.7 Flash?
Neither: both take 262K tokens.
How many benchmarks is the Qwen3.5-27B vs Step 3.7 Flash comparison based on?
8 benchmarks have a result for both models as of October 21, 2026. Qwen3.5-27B's coding index rests on 1 independent benchmark and Step 3.7 Flash's on 0. Where only one model has a result, the page says so instead of naming a winner.