Compare/GPT-5 mini (medium) vs Devstral Medium

GPT-5 mini (medium)vsDevstral Medium

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

OpenAI

GPT-5 mini (medium)

Input
$0.25/M
Output
$2/M
Speed
83 tok/s
TTFT
13.25s
Mistral

Devstral Medium

Input
$0.4/M
Output
$2/M
Speed
78 tok/s
TTFT
0.52s

Winner by Category

Cheaper
GPT-5 mini (medium)
Faster (tok/s)
GPT-5 mini (medium)
Lower Latency
Devstral Medium
Benchmarks (11-1)
GPT-5 mini (medium)

Pricing Comparison

MetricGPT-5 mini (medium)Devstral Medium
Input ($/M tokens)$0.25$0.4
Output ($/M tokens)$2$2
Cost for 1M input + 100K output tokens:
GPT-5 mini (medium)$0.45
Devstral Medium$0.60

Speed Comparison

Output Speed (tokens/s) — higher is better
GPT-5 mini (medium)
83 tok/s
Devstral Medium
78 tok/s
Time to First Token (seconds) — lower is better
GPT-5 mini (medium)
13.25s
Devstral Medium
0.52s

Benchmark Comparison

Data from Artificial Analysis API — 12 benchmarks

Intelligence Index
38.918.7
Coding Index
32.815.9
Math Index
85.04.7
GPQA Diamond
80.3%49.2%
MMLU-Pro
82.8%70.8%
LiveCodeBench
69.2%33.7%
AIME 2025
85.0%4.7%
MATH-500
70.7%
Humanity's Last Exam
14.6%3.8%
SciCode
41.0%29.4%
IFBench
71.2%29.9%
TerminalBench
28.8%9.1%
GPT-5 mini (medium)11 wins
1 winsDevstral Medium

Frequently Asked Questions

Which is cheaper, GPT-5 mini (medium) or Devstral Medium?

GPT-5 mini (medium) 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?

GPT-5 mini (medium) wins 11 out of 12 benchmarks compared to 1 for Devstral Medium. See the detailed benchmark chart above for per-category results.

Which is faster for real-time applications?

GPT-5 mini (medium) generates tokens faster at 83 tok/s vs 78 tok/s. However, Devstral Medium has lower time-to-first-token (0.52s vs 13.25s).

When should I use GPT-5 mini (medium) vs Devstral Medium?

Choose based on your priorities: GPT-5 mini (medium) for lower cost, GPT-5 mini (medium) for stronger benchmark performance, and GPT-5 mini (medium) for faster generation. For latency-sensitive apps, check the TTFT comparison above.