Mistral AI

Mixtral 8x22B

High-performance Mixture-of-Experts model.

Strong performanceEfficient inference (MoE)Good multilingual capabilitiesOpen-weights available
Today's score
86.0
Try Mixtral 8x22B

Where it ranks today

Best for / Not great for

Best for
  • Complex text generation
  • Code completion and analysis
  • Translation tasks
  • Research and fine-tuning
Not great for
  • Extremely long context summarization
  • Real-time multimodal interaction
  • Simple, low-complexity tasks where smaller models suffice

Why it ranks here

Mixtral 8x22B, as an open-weights model, offers competitive performance, particularly for complex tasks. Its Mixture-of-Experts architecture provides efficiency, making it a compelling option for developers seeking power and flexibility.

30-day trend

Score breakdown

Search trends87
Benchmarks85
Developer buzz86
News mentions86

Pricing

API: $0.00 in · $0.00 out per 1M tokens · Consumer: $0.00/mo

Pricing plans

Popular
La Plateforme (Mixtral)
Access powerful open models via API.
$5.60 /usage
  • Pay-as-you-go pricing
  • Access to Mixtral 8x22B
  • High throughput
  • Optimized inference
Use the API
Open Source Model
Deploy the model yourself.
Free
  • Downloadable weights
  • Full customization
  • Self-hosting required
  • Community support
Get the Model
Compare with another modelHow is this score calculated? →Snapshot 2026-08-15