Compare/Qwen3 30B A3B (Non-reasoning) vs Llama 3.2 Instruct 90B (Vision)

Qwen3 30B A3B (Non-reasoning)vsLlama 3.2 Instruct 90B (Vision)

Side-by-side comparison of pricing, 12 benchmarks, and generation speed.

Alibaba

Qwen3 30B A3B (Non-reasoning)

Input
$0.2/M
Output
$0.8/M
Speed
68 tok/s
TTFT
1.02s
Meta

Llama 3.2 Instruct 90B (Vision)

Input
$0.72/M
Output
$0.72/M
Speed
42 tok/s
TTFT
0.37s

Winner by Category

Cheaper
Qwen3 30B A3B (Non-reasoning)
Faster (tok/s)
Qwen3 30B A3B (Non-reasoning)
Lower Latency
Llama 3.2 Instruct 90B (Vision)
Benchmarks (11-1)
Qwen3 30B A3B (Non-reasoning)

Pricing Comparison

MetricQwen3 30B A3B (Non-reasoning)Llama 3.2 Instruct 90B (Vision)
Input ($/M tokens)$0.2$0.72
Output ($/M tokens)$0.8$0.72
Cost for 1M input + 100K output tokens:
Qwen3 30B A3B (Non-reasoning)$0.28
Llama 3.2 Instruct 90B (Vision)$0.79

Speed Comparison

Output Speed (tokens/s) — higher is better
Qwen3 30B A3B (Non-reasoning)
68 tok/s
Llama 3.2 Instruct 90B (Vision)
42 tok/s
Time to First Token (seconds) — lower is better
Qwen3 30B A3B (Non-reasoning)
1.02s
Llama 3.2 Instruct 90B (Vision)
0.37s

Benchmark Comparison

Data from Artificial Analysis API — 12 benchmarks

Intelligence Index
12.511.9
Coding Index
13.3
Math Index
21.7
GPQA Diamond
51.5%43.2%
MMLU-Pro
71.0%67.1%
LiveCodeBench
32.2%21.4%
AIME 2025
21.7%
MATH-500
86.3%62.9%
Humanity's Last Exam
4.6%4.9%
SciCode
26.4%24.0%
IFBench
31.9%
TerminalBench
6.8%
Qwen3 30B A3B (Non-reasoning)11 wins
1 winsLlama 3.2 Instruct 90B (Vision)

Frequently Asked Questions

Which is cheaper, Qwen3 30B A3B (Non-reasoning) or Llama 3.2 Instruct 90B (Vision)?

Qwen3 30B A3B (Non-reasoning) is cheaper overall. Its blended price (3:1 input/output ratio) is $0.35/M tokens vs $0.72/M for Llama 3.2 Instruct 90B (Vision).

Which model performs better on benchmarks?

Qwen3 30B A3B (Non-reasoning) wins 11 out of 12 benchmarks compared to 1 for Llama 3.2 Instruct 90B (Vision). See the detailed benchmark chart above for per-category results.

Which is faster for real-time applications?

Qwen3 30B A3B (Non-reasoning) generates tokens faster at 68 tok/s vs 42 tok/s. However, Llama 3.2 Instruct 90B (Vision) has lower time-to-first-token (0.37s vs 1.02s).

When should I use Qwen3 30B A3B (Non-reasoning) vs Llama 3.2 Instruct 90B (Vision)?

Choose based on your priorities: Qwen3 30B A3B (Non-reasoning) for lower cost, Qwen3 30B A3B (Non-reasoning) for stronger benchmark performance, and Qwen3 30B A3B (Non-reasoning) for faster generation. For latency-sensitive apps, check the TTFT comparison above.