Home/Models/Mixtral 8x7B Instruct
Mistral AI logo

Mixtral 8x7B Instruct

by Mistral AI · Released Dec 2023

Open Source
66.8
avg score
Rank #48
Compare
Better than 79% of all models
Context
33K tokens (~16 books)
Input $/1M
$0.54
Output $/1M
$0.54
Type
text
License
Open Source
Benchmarks
11 tested
Data updated today
About

Mixtral 8x7B Instruct is a pretrained generative Sparse Mixture of Experts, by Mistral AI, for chat and instruction use. Incorporates 8 experts (feed-forward networks) for a total of 47 billion...

Tested on 11 benchmarks with 57.8% average. Top scores: ARC AI2 (83.1%), HellaSwag (82.3%), TriviaQA (82.2%).

Looking for similar performance at lower cost?
GPT-5.1-Codex-Mini scores 67.4 (101% as good) at $0.25/1M input · 54% cheaper
Capabilities
math
42.2
#101 globally
knowledge
61.3
#42 globally
Benchmark Scores
Compare All
Tested on 11 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.

74.4
MATH level 5

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

9.9
ARC AI2

AI2 Reasoning Challenge. Grade-school science questions requiring multi-step reasoning. Easy and Challenge sets test different difficulty levels.

83.1
HellaSwag

Sentence completion requiring commonsense reasoning about physical and social situations. Tests real-world understanding.

82.3
TriviaQA

Trivia questions sourced from trivia enthusiasts and quiz websites. Tests breadth of general knowledge.

82.2
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Links
Documentation
Community
BenchGecko API
mixtral-8x7b-instruct
Specifications
  • Typetext
  • Context33K tokens (~16 books)
  • ReleasedDec 2023
  • LicenseOpen Source
  • StatusActive
  • Cost / Message~$0.002
Available On
Mistral AI logoMistral AI$0.54
Categories
Share & Export
Tweet
Mixtral 8x7B Instruct is an open-source text AI model by Mistral AI, released in December 2023. It has an average benchmark score of 66.8. Context window: 33K tokens.