The Naming Is Confusing

GPT-4o, Claude 3.5 Sonnet, Gemini 1.5 Pro, Llama 3.1 70B. If these names mean nothing to you, you're not alone. Here's a plain-language guide.

GPT (OpenAI)

GPT stands for "Generative Pre-trained Transformer." It's made by OpenAI, the company behind ChatGPT.

• GPT-4o: The current flagship. The "o" stands for "omni," meaning it handles text, images, and audio. Strong all-around model.
• GPT-4 Turbo: An older variant, still capable, slightly different strengths.
• GPT-3.5: The older, faster, less capable version. Used in free tiers.

Claude (Anthropic)

Claude is made by Anthropic, a company focused on AI safety. Named after Claude Shannon, the father of information theory.

• Claude 3.5 Sonnet: The balanced option. Good quality and reasonable speed.
• Claude 3 Opus: The most capable version. Slower but more thorough.
• Claude 3 Haiku: The fast, lightweight version for simple tasks.

Claude's standout feature is its massive context window (up to 200K tokens), making it excellent for long documents.

Gemini (Google)

Gemini is Google's AI model family, replacing the earlier "Bard" product.

• Gemini 1.5 Pro: Full-featured model with a very large context window (up to 1M tokens).
• Gemini 1.5 Flash: Faster, lighter version for quick tasks.
• Gemini Ultra: The most capable variant for complex reasoning.

Gemini's strength is multimodal capability and integration with Google's knowledge.

Llama (Meta)

Llama is Meta's open-source model family. "Llama" stands for "Large Language Model Meta AI."

• Llama 3.1: Available in various sizes (8B, 70B, 405B parameters).
• The "B" refers to billions of parameters. Bigger generally means more capable but slower.

Being open-source means anyone can run Llama on their own hardware, making it popular for privacy-focused and enterprise deployments.

What the Numbers Mean

• Version numbers (3, 4, 3.5): Newer is generally better
• Parameter counts (7B, 70B, 405B): More parameters = more capable but slower
• Names like Sonnet, Opus, Haiku: Quality tiers within the same model family (usually small, medium, large)

Which One Should You Pick?

Each model has strengths for different tasks. Or skip the decision entirely and let Octofy's automatic selection handle it. The platform stays current with model updates, so you're always using the best available option.