Compare · ModelsLive · 2 picked · head to head
DeepSeek V4 Flash vs GPT-5 Mini
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
DeepSeek V4 Flash wins on 8/11 benchmarks
DeepSeek V4 Flash wins 8 of 11 shared benchmarks. Leads in general · reasoning · math.
Category leads
general·DeepSeek V4 Flashcoding·GPT-5 Minireasoning·DeepSeek V4 Flashlanguage·GPT-5 Minimath·DeepSeek V4 Flashknowledge·DeepSeek V4 Flash
Hype vs Reality
Attention vs performance
DeepSeek V4 Flash
#39 by perf·#6 by attention
GPT-5 Mini
#108 by perf·#4 by attention
Best value
DeepSeek V4 Flash
2.1x better value than GPT-5 Mini
DeepSeek V4 Flash
94.4 pts/$
$0.65/M
GPT-5 Mini
45.8 pts/$
$1.13/M
Vendor risk
Mixed exposure
One or more vendors flagged
DeepSeek
$3.4B·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
11 benchmarks · 2 models
DeepSeek V4 FlashGPT-5 Mini
Dtbench
DeepSeek V4 Flash leads by +9.8
DeepSeek V4 Flash
77.3
GPT-5 Mini
67.5
LiveBench · Agentic Coding
DeepSeek V4 Flash leads by +15.0
DeepSeek V4 Flash
50.0
GPT-5 Mini
35.0
LiveBench · Coding
GPT-5 Mini leads by +6.8
DeepSeek V4 Flash
69.2
GPT-5 Mini
76.1
LiveBench · Data Analysis
DeepSeek V4 Flash leads by +18.4
DeepSeek V4 Flash
68.0
GPT-5 Mini
49.6
LiveBench · If
GPT-5 Mini leads by +1.1
DeepSeek V4 Flash
63.1
GPT-5 Mini
64.2
LiveBench · Language
DeepSeek V4 Flash leads by +1.0
DeepSeek V4 Flash
70.1
GPT-5 Mini
69.2
LiveBench · Mathematics
DeepSeek V4 Flash leads by +5.3
DeepSeek V4 Flash
79.7
GPT-5 Mini
74.4
LiveBench · Overall
DeepSeek V4 Flash leads by +6.2
DeepSeek V4 Flash
67.3
GPT-5 Mini
61.0
LiveBench · Reasoning
DeepSeek V4 Flash leads by +11.9
DeepSeek V4 Flash
70.6
GPT-5 Mini
58.6
Lmca
DeepSeek V4 Flash leads by +2.0
DeepSeek V4 Flash
42.2
GPT-5 Mini
40.2
WeirdML
GPT-5 Mini leads by +7.0
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
DeepSeek V4 Flash
45.6
GPT-5 Mini
52.7
Full benchmark table
| Benchmark | DeepSeek V4 Flash | GPT-5 Mini |
|---|---|---|
Dtbench | 77.3 | 67.5 |
LiveBench · Agentic Coding | 50.0 | 35.0 |
LiveBench · Coding | 69.2 | 76.1 |
LiveBench · Data Analysis | 68.0 | 49.6 |
LiveBench · If | 63.1 | 64.2 |
LiveBench · Language | 70.1 | 69.2 |
LiveBench · Mathematics | 79.7 | 74.4 |
LiveBench · Overall | 67.3 | 61.0 |
LiveBench · Reasoning | 70.6 | 58.6 |
Lmca | 42.2 | 40.2 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 45.6 | 52.7 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.03 | $1.28 | 1.0M tokens (~524 books) | $3.39 | |
| $0.25 | $2.00 | 400K tokens (~200 books) | $6.88 |
People also compared