Compare/Devstral Medium vs Qwen3.5 35B A3B (Non-reasoning)

Devstral MediumvsQwen3.5 35B A3B (Non-reasoning)

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

Mistral

Devstral Medium

Input
$0.4/M
Output
$2/M
Speed
128 tok/s
TTFT
0.41s
Alibaba

Qwen3.5 35B A3B (Non-reasoning)

Input
$0.25/M
Output
$2/M
Speed
119 tok/s
TTFT
0.98s

Winner by Category

Cheaper
Qwen3.5 35B A3B (Non-reasoning)
Faster (tok/s)
Devstral Medium
Lower Latency
Devstral Medium
Benchmarks (6-6)
Tie

Pricing Comparison

MetricDevstral MediumQwen3.5 35B A3B (Non-reasoning)
Input ($/M tokens)$0.4$0.25
Output ($/M tokens)$2$2
Cost for 1M input + 100K output tokens:
Devstral Medium$0.60
Qwen3.5 35B A3B (Non-reasoning)$0.45

Speed Comparison

Output Speed (tokens/s) — higher is better
Devstral Medium
128 tok/s
Qwen3.5 35B A3B (Non-reasoning)
119 tok/s
Time to First Token (seconds) — lower is better
Devstral Medium
0.41s
Qwen3.5 35B A3B (Non-reasoning)
0.98s

Benchmark Comparison

Data from Artificial Analysis API — 12 benchmarks

Intelligence Index
18.730.7
Coding Index
15.916.8
Math Index
4.7
GPQA Diamond
49.2%81.9%
MMLU-Pro
70.8%
LiveCodeBench
33.7%
AIME 2025
4.7%
MATH-500
70.7%
Humanity's Last Exam
3.8%12.8%
SciCode
29.4%29.3%
IFBench
29.9%44.5%
TerminalBench
9.1%10.6%
Devstral Medium6 wins
6 winsQwen3.5 35B A3B (Non-reasoning)

Frequently Asked Questions

Which is cheaper, Devstral Medium or Qwen3.5 35B A3B (Non-reasoning)?

Qwen3.5 35B A3B (Non-reasoning) is cheaper overall. Its blended price (3:1 input/output ratio) is $0.69/M tokens vs $0.80/M for Devstral Medium.

Which model performs better on benchmarks?

It's a tie — both models win 6 benchmarks each across 12 evaluated categories. See the detailed benchmark chart above for per-category results.

Which is faster for real-time applications?

Devstral Medium generates tokens faster at 128 tok/s vs 119 tok/s. Devstral Medium also has lower time-to-first-token (0.41s vs 0.98s).

When should I use Devstral Medium vs Qwen3.5 35B A3B (Non-reasoning)?

Choose based on your priorities: Qwen3.5 35B A3B (Non-reasoning) for lower cost, both perform similarly on benchmarks, and Devstral Medium for faster generation. For latency-sensitive apps, check the TTFT comparison above.