Getting started

What is Vinci?

What is Vinci? Vinci is an AI assistant from SimpleDirect. You can use it to ask questions, write and edit, explain ideas, brainstorm, summarize, plan, learn, and get help with code. You can chat with Vinci on the web, iOS, or Android. Developers can also use Vinci through its API or connect it to coding tools. Vinci is designed around: - Zero Data Retention by default: conversation content is not kept on Vinci's servers unless you opt in to device sync. - Canadian data residency: your account and saved data are stored in Montréal. - Open-weight models: organizations can run Vinci on their own hardware. - A consistent character: Vinci follows the same voice and principles in the app, API, and coding tools. Vinci runs on model classes — curated sets of frontier models we test and rotate behind one endpoint, so you get the strongest available without chasing model names. - Vinci Mezzo (mezzo) is fast and light for everyday chat. For code and file tasks, Forte works best. Today it runs DeepSeek V4 Flash on DeepInfra. - Vinci Forte (forte) is the default. Today it runs GLM 5.2 on DeepInfra. - Vinci Fortissimo (fortissimo) is the deepest class, for the hardest work. Today it runs Kimi K3 on Fireworks. The class name stays stable while the model behind it can rotate after evaluation. In chat, click the model badge at the top right to switch inline, or choose under Settings → Personalization. Per-app overrides are available. Read the full introduction or try Vinci on the web.

Choose where to meet Vinci

You do not need to choose one permanent home for Vinci. Think of each available surface as a different doorway into the same assistant, then start where your task already lives. Settle in on the web Choose the web app when you want room to think, write, or research. Open vinci.getsimpledirect.com in a browser on any device. You can try a few messages without creating an account. This is the broadest chat workspace: you can continue a conversation with follow-up requests, attach a document or image, and let Vinci search the web when a question needs current information. Try asking, “Compare these two options and show the trade-offs in a table,” then refine the answer from there. Keep it close on mobile The iOS and Android apps suit questions that begin away from your desk. Mobile gives you streaming chat, web search and deep research, plus copy and regenerate actions. Your sign-in is shared with the web, so you use the same Vinci account. Try mobile when a thought arrives between meetings: “Turn these rough notes into three next steps.” Bring it into the terminal If your task lives in a software project, Vinci Code is the focused choice. This beta terminal agent can read the project, edit files, run tools, and work through multi-step coding tasks. It connects to your Vinci account through browser pairing, so there is no API key to paste or model to configure. Start with the surface nearest to your work today. Tomorrow’s task may lead you through a different door.

Build custom workflows with tool calling

Use the Vinci API's tool-calling feature to build custom automation that chains multiple API calls and tool executions. This pattern lets you define tools once and let Vinci orchestrate them across multiple steps. 1. Define your tools. Create an OpenAI-compatible tools array with JSON schema definitions of what your tools do. Each tool has a name, description, and required/optional parameters. For example, a "get_weather" tool might accept a location string and return temperature. 2. Make the first request. Send your user's request with the tools array to the Vinci API. Vinci analyzes the request and decides which tools to call, if any. The response includes tool_calls that specify the tool name and arguments. 3. Execute the tool calls. Your code performs the actions: query a database, call an external API, run a script. Gather the results for each tool call. 4. Send results back. Create { "role": "tool", "tool_call_id": "...", "content": "..." } messages for each result and append them to your messages array along with Vinci's prior tool_calls. Send the updated messages back to Vinci. 5. Continue multi-step workflows. Vinci analyzes the results and may call more tools or return a final answer. Repeat steps 3–4 until Vinci returns a response without tool_calls. 6. Handle streaming if needed. Tool calling works with streaming enabled. Chunks include tool_calls deltas so your UI can show tool decisions as they arrive. Multi-turn tool use stays within your 1,000,000-token context window. Usage is billed against your shared credit balance at the rate for the class serving the request. Published API pricing per 1 million tokens is: | Class | Input | Output | |---|---:|---:| | Vinci Mezzo | $0.10 | $0.21 | | Vinci Forte (default) | $1.07 | $3.45 | | Vinci Fortissimo | $3.45 | $17.25 | Mezzo cached input is $0.02 per 1 million tokens. Mezzo is fast and light for everyday chat. For code and file tasks and this kind of tool workflow, Forte works best. Fortissimo is the deepest class, for the hardest work.

Build with the Vinci API using practical patterns

The Vinci API is OpenAI-compatible and credit-based. Here are patterns that help you ship faster and avoid common pitfalls when building applications powered by Vinci. 1. Use streaming for better UX. Set "stream": true in your request to receive Server-Sent Events of chunks as they arrive, rather than waiting for the full response. Stream the chunks to your frontend for real-time display. 2. Handle tool calling for agentic flows. Send a tools array with your request to enable function calling. Vinci returns tool_calls in the response; your code executes them and sends results back. Multi-turn tool use is supported—send prior calls and results in the messages array. 3. Understand the model routing. Pin a class with "model": "mezzo", "model": "forte", or "model": "fortissimo", or use "model": "auto" to follow the class on the account. The API default is auto, and the account default is Forte. Mezzo is fast and light for everyday chat; for code and file tasks, Forte works best. Every class accepts a 1,000,000-token context window. Published API pricing per 1 million tokens: | Class | Input | Output | |---|---:|---:| | Vinci Mezzo | $0.10 | $0.21 | | Vinci Forte (default) | $1.07 | $3.45 | | Vinci Fortissimo | $3.45 | $17.25 | Mezzo cached input is $0.02 per 1 million tokens. 4. Respect rate limits and errors. HTTP 429 (rate_limit_error) means temporary busy-time limits, not a billing problem—wait briefly and retry. HTTP 402 (budget_exhausted) means credits are exhausted. Implement exponential backoff for 429 and alert users for 402. 5. Monitor usage on the Platform. Issue scoped API keys on platform.getsimpledirect.com. The inference scope covers /chat/completions calls. Track usage on the Platform Usage page to budget and set alerts via webhooks. 6. Remember: no data retention by default. Prompts and completions are not logged or stored—only counts and metadata are recorded for billing. ZDR is automatic; opt-in features that do store data are clearly marked.

Try Vinci's capabilities in your first week

Once ordinary chat feels natural, give yourself one small Vinci experiment each day. You are not learning a control panel; you are discovering what kinds of requests can open into more capable help. Monday: ask about right now Try a current question: “What’s the weather in Ottawa this weekend?” or “What changed in the latest release?” Vinci can automatically use a live lookup or web search, and grounded answers include sources you can inspect. Tuesday: go one level deeper Bring a question that deserves research across more than one page. On the web, choose Deep research from the web menu, then ask for a comparison or a sourced overview. Vinci breaks a larger question into parts and synthesizes what it finds. Wednesday: leave room for a question Start with an open request such as, “Help me plan a workshop.” If a key detail is missing, Vinci may ask one clarifying question before it dives in. Answer in the same conversation and watch the plan take shape around your context. Thursday: let the answer do some work Ask for a calculation, a small data transformation, or a chart. Vinci Chat can run Python, JavaScript, and TypeScript/TSX in a fresh sandbox with no network access or credentials. Results, charts, and small generated files can stream back into the conversation. Friday: make one chat disappear Open a private chat for a one-off thought. Nothing from that session is persisted; when you close it, it is gone. By Friday, you will have seen a useful pattern: Vinci decides when a live tool fits, asks rather than assumes when context is missing, and can move from words to concrete output when the task calls for it.