info

llmfit info <MODEL>

# Get details about Llama 3.3 70B
llmfit info llama-3.3-70b

# Partial name match
llmfit info "llama 70b"

# Get model info as JSON
llmfit info llama-3.3-70b --json

# Process with jq
llmfit info deepseek-v3 --json | jq '.models[0].score_components'

# Check fit with 24GB VRAM
llmfit info llama-3.1-70b --memory 24G

# Check fit with 8K context limit
llmfit info qwen-2.5-72b --max-context 8192

# Compare multiple models
llmfit info llama-3.3-70b
llmfit info qwen-2.5-72b
llmfit info deepseek-v3

=== llama-3.3-70b ===

Provider: Meta
Parameters: 70B
Quantization: 4bit
Best Quant: 4bit
Context Length: 131072 tokens
Use Case: general
Category: General
Released: 2024-12-06
Runtime: MLX (baseline est. ~42.5 tok/s)

Score Breakdown:
  Overall Score: 95.2 / 100
  Quality: 95  Speed: 43  Fit: 100  Context: 100
  Baseline Est. Speed: 42.5 tok/s

Resource Requirements:
  Min VRAM: 36.4 GB
  Min RAM: 43.7 GB (CPU inference)
  Recommended RAM: 87.5 GB

Fit Analysis:
  Status: ✓ Perfect
  Run Mode: GPU
  Memory Utilization: 68.2% (43.7 / 64.0 GB)

GGUF Downloads:
  bartowski → https://huggingface.co/bartowski/Llama-3.3-70B-Instruct-GGUF
  Tip: llmfit download bartowski/Llama-3.3-70B-Instruct-GGUF --quant 4bit

Notes:
  High-quality general-purpose model from Meta
  Excellent for chat, reasoning, and code generation
  Requires significant VRAM for full GPU inference

$ llmfit info deepseek-v3

=== deepseek-v3 ===

Provider: DeepSeek
Parameters: 671B
Quantization: Q4_K_M
Best Quant: Q4_K_M
Context Length: 131072 tokens
Use Case: reasoning
Category: Reasoning
Released: 2024-12-26
Runtime: llama.cpp (baseline est. ~28.3 tok/s)

Score Breakdown:
  Overall Score: 92.1 / 100
  Quality: 98  Speed: 28  Fit: 89  Context: 100
  Baseline Est. Speed: 28.3 tok/s

Resource Requirements:
  Min VRAM: 350.2 GB
  Min RAM: 420.2 GB (CPU inference)
  Recommended RAM: 840.5 GB

MoE Architecture:
  Experts: 8 active / 256 total per token
  Active VRAM: 57.4 GB (vs 350.2 GB full model)
  Offloaded: 292.8 GB inactive experts in RAM

Fit Analysis:
  Status: ✓ Good
  Run Mode: MoE Offload
  Memory Utilization: 89.7% (57.4 / 64.0 GB)

GGUF Downloads:
  unsloth → https://huggingface.co/unsloth/DeepSeek-V3-GGUF
  Tip: llmfit download unsloth/DeepSeek-V3-GGUF --quant Q4_K_M

Notes:
  Advanced reasoning model with MoE architecture
  Sparse activation allows running on modest hardware
  GPU handles active experts, RAM holds inactive experts

$ llmfit info llama

Multiple models found. Please be more specific:
  - llama-3.3-70b
  - llama-3.1-405b
  - llama-3.1-70b
  - llama-3.1-8b
  - llama-3.2-3b
  - llama-3.2-1b
  - llama-3.2-11b-vision
  - llama-3.2-90b-vision

$ llmfit info nonexistent

No model found matching 'nonexistent'

$ llmfit info llama-3.3-70b --json

{
  "system": {
    "total_ram_gb": 64.0,
    "available_ram_gb": 58.24,
    "cpu_cores": 16,
    "gpu_vram_gb": 64.0,
    "backend": "Metal"
  },
  "models": [
    {
      "name": "llama-3.3-70b",
      "provider": "Meta",
      "parameter_count": "70B",
      "params_b": 70.0,
      "context_length": 131072,
      "use_case": "general",
      "category": "General",
      "release_date": "2024-12-06",
      "is_moe": false,
      "fit_level": "perfect",
      "run_mode": "gpu",
      "score": 95.2,
      "score_components": {
        "quality": 95.0,
        "speed": 42.5,
        "fit": 100.0,
        "context": 100.0
      },
      "estimated_tps": 42.5,
      "runtime": "MLX",
      "runtime_label": "MLX",
      "best_quant": "4bit",
      "memory_required_gb": 43.68,
      "memory_available_gb": 64.0,
      "utilization_pct": 68.2,
      "notes": [
        "High-quality general-purpose model",
        "Excellent for chat, reasoning, and code"
      ],
      "gguf_sources": [
        {
          "provider": "bartowski",
          "repo": "bartowski/Llama-3.3-70B-Instruct-GGUF"
        }
      ]
    }
  ]
}