Model Details
Qwen3.5 9B
Quick answer: Qwen3.5 9B is Alibaba's September 2025 compact model scoring 82.5% MMLU-Pro and 64.5% IFBench. Apache 2.0 — part of the Qwen3.5 compact family.
At a Glance
Where Qwen3.5 9B leads
- 82.5% MMLU-Pro — competitive for 9B scale
- 64.5% IFBench — instruction following at compact scale
- Apache 2.0 — fully open
- Minimal hardware requirements at 9B
Where it lags
- Limited benchmark coverage (2 metrics)
- Compact 9B: constrained reasoning vs larger models
Best for: Edge/embedded deployments at 9B scale; Apache 2.0 instruction following in constrained environments.
What Qwen3.5 9B Is
Qwen3.5 9B is the compact 9B model in Alibaba's September 2025 Qwen3.5 generation. The Qwen3.5 family ranges from 0.8B to 397B A17B MoE; the 9B variant targets minimal-resource deployments.
Specifications
| Field | Value |
|---|---|
| Organization | Alibaba |
| License | Apache 2.0 |
| HuggingFace | Qwen/Qwen3.5-9B |
| Release date | September 2025 |
| Parameters | 9B |
| Modality | Text only |
Pricing
Open weights under Apache 2.0 — self-host at no cost.
Public Benchmark Scores
| Benchmark | Score | Source | Date |
|---|---|---|---|
| MMLU-Pro | 82.5% | Benchgen evaluation | 2025-09 |
| IFBench | 64.5% | Benchgen evaluation | 2025-09 |
Qwen3.5 9B vs Alternatives
| Model | MMLU-Pro | IFBench | License |
|---|---|---|---|
| Qwen3.5 9B | 82.5% | 64.5% | Apache 2.0 |
| Qwen3.5 27B | — | 76.5% | Apache 2.0 |
| Qwen2.5 7B Instruct | 56.3% | — | Apache 2.0 |
For larger scale: Qwen3.5 27B or the dense 35B A3B offer better benchmarks with more compute.
Related Models & Benchmarks
- Compare: Qwen3.5 27B · Qwen3.5 35B A3B
- Benchmarks: MMLU-Pro · IFBench
Frequently Asked Questions
What is Qwen3.5 9B?
Alibaba's September 2025 9B model scoring 82.5% MMLU-Pro, 64.5% IFBench. Apache 2.0 — compact edge deployment.Specs from Alibaba's Qwen3.5 9B release (September 2025) and Benchgen evaluations. Last updated 2026-07-24.
Benchmark Leaderboards
This model isn’t on any benchmark leaderboard yet.