Home/Models/Mistral Nemo
Mistral AI logo

Mistral Nemo

by Mistral AI · Released Jul 2024

Open Source
37.4
avg score
Rank #158
Compare
Better than 32% of all models
Context
131K tokens (~66 books)
Input $/1M
$0.02
Output $/1M
$0.03
Type
text
License
Open Source
Benchmarks
5 tested
Data updated today
About

A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA. The model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese,...

Tested on 5 benchmarks with 37.2% average. Top scores: GSM8K (84.2%), PIQA (67.0%), Balrog (17.6%).

Capabilities
math
47.5
#81 globally
knowledge
30.4
#175 globally
Benchmark Scores
Compare All
Tested on 5 benchmarks · Ranked across 2 categories
Score Distribution (all 233 models)
0255075100
▲ You are here
GSM8K

Grade school math word problems. 8,500 problems testing multi-step arithmetic reasoning. A foundational math benchmark.

84.2
MATH level 5

Competition-level math from AMC, AIME, and olympiad problems. Level 5 is the hardest tier, requiring creative problem-solving.

10.8
PIQA

Physical Intuition QA. Tests understanding of everyday physical interactions and commonsense physics.

67.0
Balrog

Broad Assessment of Language and Reasoning Over Games. Tests strategic and logical reasoning through game scenarios.

17.6
GPQA diamond

Graduate-level science questions written by PhD experts. Diamond subset contains questions where experts disagree, testing deep understanding.

6.5
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Links
Documentation
Community
BenchGecko API
mistral-nemo
Specifications
  • Typetext
  • Context131K tokens (~66 books)
  • ReleasedJul 2024
  • LicenseOpen Source
  • StatusActive
  • Cost / Message~$0.000
Available On
Mistral AI logoMistral AI$0.02
Categories
Share & Export
Tweet
Mistral Nemo is an open-source text AI model by Mistral AI, released in July 2024. It has an average benchmark score of 37.4. Context window: 131K tokens.