BenchMark'd
Based on 1 review
A dense 27.8B model from the Qwen 3.5 Small series, optimized for on-device and cost-efficient deployment. Strong reasoning and coding performance with 256K context window support.
Released
March 2, 2026
Parameters
27.8B
Context
256K
Pricing
Paid
| Benchmark | Category | Score | Performance |
|---|---|---|---|
MMLU | knowledge | 85.7% | 86 |
HumanEval | coding | 82.3% | 82 |
MATH | reasoning | 80.9% | 81 |
Chatbot Arena ELO | overall | 1345 | 78 |
Last updated: March 15, 2026
Benchmark scores may vary based on evaluation methodology and conditions.