Model comparison
Qwen3.5-9B vs Qwen3.8 27B
Qwen3.8 27B is the stronger model overall, scoring 46.0 to 33.8 on the Noometry Index. Qwen3.5-9B costs 9.9× less per token, which makes it the better buy when Qwen3.8 27B's lead doesn't matter for your workload.
Last verified . 5 shared benchmarks.
Alibaba (Qwen)Summary
- They share 5 benchmarks with published results for both. Qwen3.5-9B scores higher in 1 category and Qwen3.8 27B in 4 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Qwen3.8 27B leads 32.9 to 14.5.
- The biggest single-benchmark swing is SciCode: 27.5% for Qwen3.5-9B and 46.6% for Qwen3.8 27B.
- Qwen3.5-9B is cheaper at $0.10 / $0.15 per million input/output tokens, against $0.99 / $1.49 for Qwen3.8 27B.
Side by side
| Qwen3.5-9B | Qwen3.8 27B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 33.8 | 46.0 |
| Released | 2026-02-23 | 2026-08-14 |
| Weights | Open | Open |
| Context window | 262K | 262K |
| Max output | 66K | 33K |
| Input $ / M tokens | $0.10 | $0.99 |
| Output $ / M tokens | $0.15 | $1.49 |
| Results tracked | 10 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 27B leads
Qwen3.5-9B: 35.9 (#217), Qwen3.8 27B: 50.5 (#44)
| Benchmark | Qwen3.5-9B | Qwen3.8 27B |
|---|---|---|
| SciCode | 27.5% | 46.6% |
| LMArena WebDev | — | 1593 |
| LMArena Coding | — | 1482 |
Agentic & Tool Use Qwen3.8 27B leads
Qwen3.5-9B: 14.5 (#151), Qwen3.8 27B: 32.9 (#57)
| Benchmark | Qwen3.5-9B | Qwen3.8 27B |
|---|---|---|
| Terminal-Bench | 9.2% | — |
| APEX-Agents | — | 47.5% |
Reasoning Qwen3.8 27B leads
Qwen3.5-9B: 23.1 (#182), Qwen3.8 27B: 41.0 (#54)
| Benchmark | Qwen3.5-9B | Qwen3.8 27B |
|---|---|---|
| CritPt | 0.3% | 5.4% |
| DTBench | 71.2% | 88% |
| LMCA | 24.5% | 41.4% |
| Epoch Capabilities Index | 139.46 | 149.38 |
| ARC-AGI-2 | — | 42.4% |
| NYT Connections (extended) | — | 54.5% |
| ARC-AGI-1 | — | 87.5% |
| Chess Puzzles | 12% | — |
| LMArena Hard Prompts | — | 1460 |
| Surface Evolver Bench | — | 45% |
Math Qwen3.8 27B leads
Qwen3.5-9B: 34.8 (#192), Qwen3.8 27B: 37.1 (#161)
| Benchmark | Qwen3.5-9B | Qwen3.8 27B |
|---|---|---|
| MathArena Final-Answer Competitions | 48.5% | — |
| OTIS Mock AIME 2024-2025 | 61.7% | — |
| ProofBench | — | 16% |
| LMArena Math | — | 1456 |
Knowledge Qwen3.5-9B leads
Qwen3.5-9B: 46.0 (#84), Qwen3.8 27B: 41.6 (#109)
| Benchmark | Qwen3.5-9B | Qwen3.8 27B |
|---|---|---|
| GPQA Diamond | 79% | — |
| LMArena Expert | — | 1482 |
Multimodal Not comparable
Qwen3.5-9B: —, Qwen3.8 27B: 41.3 (#37)
| Benchmark | Qwen3.5-9B | Qwen3.8 27B |
|---|---|---|
| LMArena Vision | — | 1271 |
Multilingual Not comparable
Qwen3.5-9B: —, Qwen3.8 27B: 53.7 (#60)
| Benchmark | Qwen3.5-9B | Qwen3.8 27B |
|---|---|---|
| LMArena Non-English | — | 1430 |
| LMArena Chinese | — | 1504 |
| LMArena French | — | 1465 |
| LMArena German | — | 1438 |
| LMArena Japanese | — | 1384 |
| LMArena Korean | — | 1393 |
| LMArena Russian | — | 1415 |
| LMArena Spanish | — | 1448 |
Instruction Following Not comparable
Qwen3.5-9B: —, Qwen3.8 27B: 75.8 (#53)
| Benchmark | Qwen3.5-9B | Qwen3.8 27B |
|---|---|---|
| LMArena Instruction Following | — | 1439 |
Long Context Not comparable
Qwen3.5-9B: —, Qwen3.8 27B: 44.3 (#70)
| Benchmark | Qwen3.5-9B | Qwen3.8 27B |
|---|---|---|
| LMArena Longer Query | — | 1450 |
Writing & Preference Not comparable
Qwen3.5-9B: —, Qwen3.8 27B: 65.8 (#43)
| Benchmark | Qwen3.5-9B | Qwen3.8 27B |
|---|---|---|
| LMArena Text | — | 1441 |
| LMArena Creative Writing | — | 1384 |
| EQ-Bench Creative Writing | — | 1671 |
| LMArena Multi-Turn | — | 1441 |
Frequently asked questions
Is Qwen3.5-9B better than Qwen3.8 27B?
Qwen3.8 27B is the stronger model overall, scoring 46.0 to 33.8 on the Noometry Index. Qwen3.5-9B costs 9.9× less per token, which makes it the better buy when Qwen3.8 27B's lead doesn't matter for your workload.
Which is cheaper, Qwen3.5-9B or Qwen3.8 27B?
Qwen3.5-9B is cheaper. It lists at $0.10 per million input tokens and $0.15 per million output tokens; Qwen3.8 27B lists at $0.99 and $1.49.
Is Qwen3.5-9B or Qwen3.8 27B better for coding?
Qwen3.8 27B scores higher on coding benchmarks: 50.5 versus 35.9 in the Noometry coding category.
Which has the bigger context window?
Both accept 262K tokens.
How many benchmarks do Qwen3.5-9B and Qwen3.8 27B share?
5 benchmarks have published results for both models. Qwen3.5-9B has 10 scored results on Noometry and Qwen3.8 27B has 31.