Qwen3.6 35B A3B (Non-reasoning)
Analysis Summary
Qwen3.6 35B A3B (Non-reasoning) is Alibaba's mixture-of-experts variant at the 35B active parameter scale, running in non-reasoning mode. Its agentic index of 55.4 and GPQA of 0.817 are strong for the price point, and its long-context reliability score of 0.567 suggests reasonable performance on extended documents. The coding index of 17.6 is a notable gap relative to its other strengths.
For businesses, this model suits content generation, instruction-following pipelines, and tasks requiring strong agentic coordination without heavy coding demands. Its pricing at $0.375 input is competitive for the capability level. The non-reasoning mode limits its utility on tasks requiring deliberate multi-step analysis.
A -4 point regional penalty applies given the provider's limited enterprise footprint. Teams comfortable with Alibaba's API infrastructure will find it a cost-efficient option for content and agentic workflows.
Assessed June 6, 2026
Editorial notes
Qwen3.6 35B A3B (Non-reasoning) from Alibaba offers a strong agentic index of 55.4 and good instruction following at $0.375/$2.25 per million tokens, though its coding index is limited.
Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?
Performance Profile
How Qwen3.6 35B A3B (Non-reasoning) compares
Qwen3.6 35B A3B (Non-reasoning) ranks #99 of 378 AI models we track for overall intelligence, #159 of 315 for coding, #67 of 289 for agentic tasks. At $0.38 per million input tokens it is cheaper than 39% of comparable models.
Performance Indices
Source: Artificial Analysis
Benchmark Scores
Intelligence
Technical
Content
Benchmark data from Artificial Analysis and Hugging Face
How does Qwen3.6 35B A3B (Non-reasoning) stack up?
Compare side-by-side with other professional models.
Model Information
| Provider | Alibaba |
| Release Date | April 16, 2026 |
| Status | Active |
Pricing
| Token Type | Cost per 1M tokens | Cost per 1K tokens |
|---|---|---|
| Input | $0.38 | $0.000375 |
| Output | $2.25 | $0.002250 |
Leaderboard Categories
Frequently asked questions about Qwen3.6 35B A3B (Non-reasoning)
How much does Qwen3.6 35B A3B (Non-reasoning) cost?
Qwen3.6 35B A3B (Non-reasoning) costs $0.38 per million input tokens and $2.25 per million output tokens.
Is Qwen3.6 35B A3B (Non-reasoning) good for coding?
On our coding benchmark index, Qwen3.6 35B A3B (Non-reasoning) ranks #159 of 315 models, placing it in the broader range of the field for code generation and debugging.
Who created Qwen3.6 35B A3B (Non-reasoning)?
Qwen3.6 35B A3B (Non-reasoning) is developed by Alibaba and was released on April 16, 2026.
