Documentation

🔍Model Discovery & Search

Access 400+ models from 60+ providers via OpenRouter with consensus validation

← Back to Documentation

🔍 Model Discovery & Search

Access 400+ AI models from 60+ providers through OpenRouter

What Makes Our Model Discovery Unique?

  • 🏆 Complete Provider Independence - Provider-agnostic discovery across ALL major AI companies
  • 🌐 Universal AI Model Access - Think of it as the "Google for AI models"
  • ⚡ Real-time Intelligence - Live performance scoring and cost comparison
  • 🔄 Daily Updates - Never miss new releases or price changes

🚀 Powerful CLI Commands

# Find models good at coding
hive models search "coding"
→ Returns 20 models ranked by programming performance

# Find budget-friendly options
hive models search "budget friendly"
→ Smart cost analysis across all providers

# Side-by-side model comparison
hive models compare gpt-4 claude-3-opus
→ Detailed performance metrics, costs, and winner analysis

# Search by provider or capability
hive models search "anthropic"
→ Fuzzy matching across model names, descriptions, and IDs

🔥 Live OpenRouter Intelligence: 400+ Models from 60+ Providers

Major Providers Include:

  • OpenAI (39 models)
  • Anthropic (24 models)
  • Google (23 models)
  • Meta (26 models)
  • Plus 56+ more providers

🎯 Smart Ranking Features:

  • Programming Leaderboards - Live OpenRouter rankings
  • Usage Analytics - Real popularity percentages
  • Cost Intelligence - Per-million token pricing
  • Fuzzy Search - Find models by any keyword
  • Instant Comparisons - Side-by-side performance tables

🎯 Seamless Consensus Integration

Discovered models integrate directly with our 4-stage consensus pipeline:

From Discovery to Consensus:

# Find the best coding model
hive models search "coding"

# Use it in consensus pipeline
hive consensus --generator "anthropic/claude-3-sonnet"

🔄 4-Stage Pipeline Benefits:

  • Generator → Initial response creation
  • Refiner → Enhancement and improvement
  • Validator → Quality assurance check
  • Curator → Final polishing pass

Model Capabilities Scoring

Every model gets scored 0-100 in different categories:

🧠 Reasoning (0-100)

Logic puzzles and problem-solving

💻 Coding (0-100)

Code generation and debugging

✍️ Writing (0-100)

Content quality and creativity

🔢 Math (0-100)

Arithmetic and formula derivation

🎨 Creativity (0-100)

Original ideas and content

💰 Value Score (0-100)

Quality per dollar spent

Ready to Access Every AI Model on Earth?

Break free from vendor lock-in with the only independent AI model discovery platform.

# Install the platform
npm install -g @hivetechs/hive-ai

# Discover models perfect for your needs
hive models search "what you want to accomplish"
Start free trial