Pick from 307+ language and reasoning models across 37+ providers, ground rows with 5 native Live Web Search engines, or gate lists with calibrated Decision Mode inside Claude MCP, ChatGPT, Cursor, and Google Sheets.
307+ language & reasoning models · 5 live search engines · Calibrated Decision Mode.
OpenAI · GPT-4o · o3-mini
Structured JSON extraction, tool calling, and grounded web search.
Hover any mode to inspect its execution profile, or click to jump straight into the live directory filtered to that mode.
Filter by intelligence mode, model provider, context window, or capability. Copy any model ID for Claude MCP, ChatGPT, Cursor, or Google Sheets.
Loading the live model directory...
Compare context windows, structured output support, live web search grounding, and calibrated evaluation across leading model providers.
| Provider family | Intelligence modes | Flagship models | Context & strengths |
|---|---|---|---|
| OpenAI | Generative, Reasoning & Search | GPT-4o, GPT-4o mini, o3-mini | 128K+ ctx · JSON schemas, Vision, Tool Use & Web Search |
| Anthropic | Generative, Reasoning & Search | Claude Sonnet 4.5, Claude Haiku 4.5 | 200K ctx · High-craft outbound copy, Vision & Tool Use |
| Google Gemini | Generative, Long Context & Search | Gemini 2.5 Pro, Gemini 2.5 Flash | Up to 2M ctx · Multi-page synthesis, Vision & Web Grounding |
| Perplexity | Live Web Search | Sonar, Sonar Pro, Sonar Reasoning | 128K ctx · Native real-time web citations & news research |
| DeepSeek & Mistral | Reasoning & Fast Extraction | DeepSeek R1, DeepSeek V3, Mistral Large | 128K ctx · Chain-of-thought reasoning & multilingual JSON |
| xAI, Meta & Qwen | Open & Frontier Families | Grok, Llama 3.3, Qwen 2.5, Kimi | 128K+ ctx · Live search on xAI plus broad open-weights choice |
| JEV | Calibrated Decision Mode | JEV | 64K ctx · Calibrated Keep, Unsure & Reject probabilities |
Every model in this directory is immediately available across all five Sorzio surfaces with one shared catalog and encrypted key vault.
FAQ
Direct answers on our 307+ model catalog, 5 Live Web Search engines, calibrated Decision Mode, and encrypted API keys.
Sorzio provides instant access to 307+ language and reasoning models across 37+ providers, including OpenAI, Anthropic Claude, Google Gemini, Perplexity Sonar, Mistral, DeepSeek, xAI Grok, Meta Llama, Cohere, NVIDIA, Amazon Nova, Alibaba Qwen, and Moonshot Kimi.
Generative AI drafts 1-to-1 outbound copy, summarizes research, and extracts structured JSON. Live Web Search grounds responses in real-time web citations across 5 native search providers. Decision Mode uses JEV to output calibrated Keep, Unsure, and Reject probabilities without chat filler.
Yes. You can gate prospects first with Decision Mode, research passing accounts with Perplexity Sonar or Google Gemini web search, and write tailored cold openers with Claude or GPT-4o inside one Flow or MCP session.
Eligible workspaces can connect encrypted OpenAI, Anthropic, Google Gemini, and other provider keys using context-bound AES-256-GCM encryption. Sorzio sends zero-training flags to model providers so customer prompts and prospect rows are never used to train models.
Switch between OpenAI, Anthropic, Gemini, Perplexity Sonar, DeepSeek, and calibrated Decision Mode inside the tools your team already uses.