list

List all available LLM models

Synopsis

llmfit list

Description

Displays all LLM models in the llmfit database without filtering by system compatibility. This shows the complete catalog of models that llmfit knows about.

The list includes models from various providers (Meta, Anthropic, OpenAI, Google, Alibaba, etc.) with their parameter counts, quantizations, and context windows.

Options

This command has no options.

Usage Examples

List All Models

# Display complete model catalog
llmfit list

Count Models

# Count total models in database
llmfit list | grep -c "│"

Filter by Provider

# Find all Meta models
llmfit list | grep Meta

# Find all Anthropic models
llmfit list | grep Anthropic

Example Output

=== Available LLM Models ===
Total models: 159

╭─────────┬──────────────────────────┬───────────┬──────┬────────┬──────────────┬──────────┬────────────┬─────────┬────────┬─────────╮
│ Status  │ Model                    │ Provider  │ Size │ Score  │ tok/s est.   │ Quant    │ Runtime    │ Mode    │ Mem %  │ Context │
├─────────┼──────────────────────────┼───────────┼──────┼────────┼──────────────┼──────────┼────────────┼─────────┼────────┼─────────┤
│ --      │ llama-3.3-70b           │ Meta      │ 70B  │ -      │ -            │ 4bit     │ -          │ -       │ -      │ 128k    │
│ --      │ llama-3.1-405b          │ Meta      │ 405B │ -      │ -            │ Q4_K_M   │ -          │ -       │ -      │ 128k    │
│ --      │ qwen-2.5-72b            │ Alibaba   │ 72B  │ -      │ -            │ 4bit     │ -          │ -       │ -      │ 32k     │
│ --      │ deepseek-v3             │ DeepSeek  │ 671B │ -      │ -            │ Q4_K_M   │ -          │ -       │ -      │ 128k    │
│ --      │ claude-3.5-sonnet       │ Anthropic │ -    │ -      │ -            │ -        │ -          │ -       │ -      │ 200k    │
│ --      │ gpt-4o                  │ OpenAI    │ -    │ -      │ -            │ -        │ -          │ -       │ -      │ 128k    │
│ --      │ gemini-2.0-flash        │ Google    │ -    │ -      │ -            │ -        │ -          │ -       │ -      │ 1000k   │
│ --      │ mistral-large-2         │ Mistral   │ 123B │ -      │ -            │ Q4_K_M   │ -          │ -       │ -      │ 128k    │
│ --      │ phi-4                   │ Microsoft │ 14B  │ -      │ -            │ Q4_K_M   │ -          │ -       │ -      │ 16k     │
│ --      │ codestral-25.01         │ Mistral   │ 22B  │ -      │ -            │ Q4_K_M   │ -          │ -       │ -      │ 256k    │
╰─────────┴──────────────────────────┴───────────┴──────┴────────┴──────────────┴──────────┴────────────┴─────────┴────────┴─────────╯

Model Database

The model database is:

• Embedded: Compiled into the binary from data/hf_models.json
• No network: All data is local, no API calls required
• Updated: Regenerate via python3 scripts/scrape_hf_models.py

Model Fields

• Name: Model identifier
• Provider: Company/organization (Meta, Anthropic, etc.)
• Size: Parameter count (e.g., 7B, 70B, 405B)
• Quant: Quantization format (4bit, Q4_K_M, Q8_0, etc.)
• Context: Context window in tokens (k = thousands)

Model Types

• Local models: Have parameter counts and can run locally (llama.cpp, MLX)
• API-only models: Claude, GPT-4, Gemini (no local inference)
• Quantized: GGUF models for llama.cpp
• MLX models: 4bit quantized for Apple Silicon

Comparison with Other Commands

Command	Purpose	Filtering
list	Show all models	None
fit	Show compatible models	By system specs
search	Find specific models	By search query
recommend	Get top recommendations	By fit quality + filters

Related Commands

• search - Search for specific models
• fit - Find compatible models
• info - Get detailed model information
• recommend - Get top recommendations