← Back to all models
M

Ministral 3 14B

by Mistral AI
Efficient Free & Open Source 🏆 Ranked #64 of 85
64.0
Overall Score
out of 100
About

Mistral AI's 14B model from the Ministral 3 series, offering strong performance at a size that fits comfortably on a single 16GB GPU with quantisation.

Key Metrics
Context Window
128K
tokens
Avg Response
550
milliseconds
Input Cost
$0.08
per million tokens
Output Cost
$0.08
per million tokens
Arena ELO
1185
Chatbot Arena rating
MT-Bench
8.2
out of 10
Benchmark Scores
MMLU
73.0%
HumanEval
78.0%
MATH
62.0%
GPQA
35.0%
MT-Bench
82.0/10
Capability Profile
Strengths & Limitations
Strengths
✓ Good size-to-quality ratio ✓ Open source ✓ Fast ✓ Strong instruction following ✓ Large context
Limitations
⚠ Below 70B class ⚠ Less capable than Nemo for multilingual
Ideal Use Cases
Research Chatbots Summarisation Code assistance Business applications
Model Details
Provider Mistral AI
Released 2025-04-01
Type Free & Open Source
Multimodal No
Tier Efficient
Global rank #64 / 85
Pricing (USD)
Input tokens $0.08/M
Output tokens $0.08/M
Per 1,000 tokens ≈ $0.0001 input / $0.0001 output
All Benchmarks
MMLU 73.0%
HumanEval 78.0%
MATH 62.0%
GPQA 35.0%
MT-Bench 8.2/10
Arena ELO 1185
Compare this model View Rankings

You might also consider