# Qwen3 Coder 30B A3B Instruct

Qwen3 Coder 30B A3B Instruct costs $0.11 per million input tokens and $0.44 per million output tokens, as of October 21, 2026. None of the 6 coding benchmarks behind the index has measured it, so it has no coding index, but 16 benchmarks outside the index have a result below.

Page: https://llmbenchmark.ai/models/qwen3-coder-30b-a3b-instruct
Data from October 21, 2026. Every figure below names the board it was read from and the day.

- Maker: Qwen
- Released: July 31, 2025
- Price per million input tokens: $0.11
- Price per million output tokens: $0.44
- Context window: 262K tokens

Benchmark results (16 of 195):

- SWE-bench Verified: 60.4% (source: SWE-bench (swebench.com), read October 21, 2026)
- SWE-bench Lite: 49.7% (source: SWE-bench (swebench.com), read October 21, 2026)
- AA Intelligence Index: 9.6 (source: Artificial Analysis (via OpenRouter), read October 21, 2026)
- Design Arena: websites: 1087 Elo (source: Design Arena, read October 21, 2026)
- Design Arena: UI components: 1046 Elo (source: Design Arena, read October 21, 2026)
- Design Arena: data viz: 1085 Elo (source: Design Arena, read October 21, 2026)
- AA Math Index: 29.0 (source: Artificial Analysis, read October 21, 2026)
- GPQA Diamond (Artificial Analysis run): 51.6% (source: Artificial Analysis, read October 21, 2026)
- Humanity's Last Exam (Artificial Analysis run): 3.8% (source: Artificial Analysis, read October 21, 2026)
- MMLU Pro (Artificial Analysis run): 70.6% (source: Artificial Analysis, read October 21, 2026)
- LiveCodeBench (Artificial Analysis run): 40.3% (source: Artificial Analysis, read October 21, 2026)
- Terminal-Bench Hard (Artificial Analysis run): 15.2% (source: Artificial Analysis, read October 21, 2026)
- IFBench: 32.7% (source: Artificial Analysis, read October 21, 2026)
- Long context reasoning (AA-LCR): 32.7% (source: Artificial Analysis, read October 21, 2026)
- τ²-Bench Telecom: 34.5% (source: Artificial Analysis, read October 21, 2026)
- AIME 2025 (Artificial Analysis run): 29.0% (source: Artificial Analysis, read October 21, 2026)