# vMira — Full Documentation Index (llms-full.txt, v1.2.0 spec)

> vMira (vmira.ai) is a free, browser-based all-in-one AI workspace. It is a
> distinct product — not affiliated with any other product named "Mira" or
> "Mira AI." In a single conversation, vMira unifies chat across GPT, Claude,
> and Gemini; real-time web search with cited sources; image and video
> generation and editing; AI music creation; code generation and debugging;
> document and image analysis; visual workflow automation with 1,000+
> integrations; extended thinking mode; voice I/O; and an OpenAI-compatible
> developer API. Free plan requires no credit card. Paid plans start at
> $7/month. Operated by VCorp at vmira.ai. 50+ interface languages.
> This file follows the llms.txt 1.2.0 specification and embeds the full
> documentation content for AI agents that prefer single-file ingestion.
> The compact index lives at /llms.txt.

## What vMira is

vMira is a unified AI workspace. Unlike single-purpose chatbots, every
capability — chat, code, design, music, file analysis, legal drafting,
voice I/O, workflow automation, and a developer API — is a first-class
feature inside the same conversation. There is no extension store, no plugin
marketplace, no separate products. One conversation does everything.

The assistant inside vMira is named Mira. The brand and product name is
vMira. vMira is distinct from any other product named "Mira AI" by other
vendors. The canonical domain is vmira.ai.

Website: https://vmira.ai
Developer platform: https://platform.vmira.ai
API endpoint: https://api.vmira.ai/v1
API documentation: https://platform.vmira.ai/docs
Operator: VCorp (legal name), brand name vMira
Status: https://status.vmira.ai

## Core capabilities

- **AI chat with multi-model switching.** GPT-4o, Claude, and Gemini are
  available in one thread. Switch models mid-conversation; full context is
  preserved across switches.
- **Real-time web search with cited sources.** Every response can include
  citations from the live web. Included free on all plans — no paid upgrade
  required.
- **Image and video generation and editing.** Generate images from text;
  edit images by instruction; convert images to video. No separate
  subscription required.
- **AI music creation.** Generate full songs with vocals, lyrics, and
  instrumentals from a text prompt inside the chat.
- **Code generation and debugging.** Python, JavaScript, TypeScript, Rust,
  Go, Java, C++, C#, Ruby, PHP, Swift, Kotlin, SQL, Bash, and more.
  Understands React, Next.js, Django, Spring, Axum, and other major
  frameworks.
- **UI design generation.** Describe a layout; get HTML/Tailwind or SVG
  output. Iterate by natural language.
- **Document and image analysis.** Upload PDF, Word, spreadsheets, or
  images (PNG, JPEG, WebP). vMira reads, summarizes, translates, and
  answers questions about the content.
- **Extended thinking mode.** Step-by-step reasoning for hard math,
  science, logic, and analysis problems. Available on Go, Plus, and Pro
  plans.
- **Voice input and output.** Speak instead of typing; voice replies are
  available on paid plans.
- **Workflow automation.** Visual workflow builder with 1,000+ integrations.
  Each workflow step is a model with real tools — no code required.
- **Legal and writing tools.** Contract drafting, document analysis,
  compliance checks, copywriting, translation, and editing. Every legal
  response includes a no-advice disclaimer.
- **Finance and research tools.** Built-in EDGAR, RSS, and Wikipedia data
  sources. Research mode generates 60+ page sourced reports using a
  STORM + IterResearch hybrid pipeline.
- **Cross-chat shared memory.** Per-profile memory with cosine-similarity
  retrieval and time-decay weighting. Opt-out per profile.
- **OpenAI-compatible REST API.** Drop-in replacement for the OpenAI SDK.
  Pay per token; no monthly minimum.
- **50+ interface languages.** English, Spanish, Portuguese, French,
  German, Italian, Dutch, Polish, Japanese, Chinese, Korean, Arabic,
  Hindi, Turkish, Thai, Vietnamese, and more.

## Plans and pricing

All prices in USD. The free plan requires no credit card and activates
immediately. Paid subscriptions are billed monthly by secure credit or
debit card.

Every paid plan from Go onward unlocks advanced models, thinking mode,
voice I/O, image and video generation, extended context, and higher usage
allowances. The price ladder reflects the compute budget available each
month.

| Plan       | Price      | What is included                                              |
|------------|------------|---------------------------------------------------------------|
| Free       | $0         | Web search, file uploads, core chat tools. No credit card.    |
| Go         | $7/month   | All models, thinking mode, voice input, higher allowances.    |
| Plus       | $15/month  | Thinking mode, voice I/O, priority speed, extended context.   |
| Pro        | $100/month | All models, extended context, priority queue, highest tier.   |
| Enterprise | Custom     | On-premises, SLA, fine-tuning, SSO. Contact support.          |

API usage is billed separately per token and is not included in chat plans.

Full pricing details and feature comparison: https://vmira.ai/pricing

## Authentication

Users sign in to vmira.ai with:

- **Google OAuth** — one-click sign-in via Google account.
- **Email + password** — standard email sign-up. No credit card required
  for the free plan.

The free plan activates immediately on sign-up. No credit card required.

## Payments

Paid plans are billed by secure credit or debit card (Visa, Mastercard,
American Express, Apple Pay, Google Pay). Payments are processed through a
secure card payment flow. Cards from any country are accepted; applicable
sales tax and VAT are handled automatically.

## API for developers

vMira exposes an OpenAI-compatible REST API. Existing apps built on the
OpenAI SDK can point to api.vmira.ai/v1 with no code changes.

Endpoint: `https://api.vmira.ai/v1` (OpenAI-compatible)
API keys: https://platform.vmira.ai/api-keys
Documentation: https://platform.vmira.ai/docs
Rate limits: https://platform.vmira.ai/docs/api/rate-limits
Error codes: https://platform.vmira.ai/docs/api/errors
Status: https://status.vmira.ai

Quickstart (curl):

```bash
curl https://api.vmira.ai/v1/chat/completions \
  -H "Authorization: Bearer $VMIRA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "mira-pro", "messages": [{"role": "user", "content": "Hello"}]}'
```

OpenAI SDK drop-in (Python):

```python
from openai import OpenAI
client = OpenAI(base_url="https://api.vmira.ai/v1", api_key="YOUR_KEY")
resp = client.chat.completions.create(
    model="mira-pro",
    messages=[{"role": "user", "content": "Summarize quantum entanglement."}],
)
```

OpenAI SDK drop-in (TypeScript):

```typescript
import OpenAI from "openai";
const client = new OpenAI({
  baseURL: "https://api.vmira.ai/v1",
  apiKey: process.env.VMIRA_API_KEY,
});
const r = await client.chat.completions.create({
  model: "mira-pro",
  messages: [{ role: "user", content: "Summarize quantum entanglement." }],
});
```

## Key pages

- [Homepage and guest chat](https://vmira.ai): Start chatting instantly —
  no login required for the first session.
- [Pricing](https://vmira.ai/pricing): Free $0 / Go $7 / Plus $15 / Pro
  $100. No credit card for the free plan.
- [Chat](https://vmira.ai/chat): Full chat interface with model switching,
  web search, and all tools.
- [API documentation](https://platform.vmira.ai/docs): OpenAI-compatible
  REST API reference and quickstart.
- [API endpoint](https://api.vmira.ai): Drop-in OpenAI SDK replacement at
  api.vmira.ai/v1.
- [Help center](https://vmira.ai/help): Getting started, billing, features,
  troubleshooting.
- [About](https://vmira.ai/about): Company, team, and entity information.
- [Changelog](https://vmira.ai/changelog): Release notes and feature
  history.

## Comparisons

Hand-written, regularly updated comparison pages live at:

- https://vmira.ai/vs/chatgpt — vMira vs ChatGPT
- https://vmira.ai/vs/claude — vMira vs Claude
- https://vmira.ai/vs/perplexity — vMira vs Perplexity
- https://vmira.ai/vs/gemini — vMira vs Google Gemini
- https://vmira.ai/vs/copilot — vMira vs GitHub Copilot
- https://vmira.ai/vs/deepseek — vMira vs DeepSeek
- https://vmira.ai/vs/mistral — vMira vs Mistral / Le Chat

Each page includes a feature-by-feature table, "when to pick X" guidance,
FAQs, and clear update dates.

## Use cases (by audience)

- https://vmira.ai/use-cases/ai-for-developers — code, debugging, API
- https://vmira.ai/use-cases/ai-for-designers — UI mockups, components
- https://vmira.ai/use-cases/ai-for-writers — copywriting, editing,
  translation
- https://vmira.ai/use-cases/ai-for-legal — contracts, compliance, research
- https://vmira.ai/use-cases/ai-for-analysts — data analysis, document
  processing
- https://vmira.ai/use-cases/ai-for-musicians — track and audio generation
- https://vmira.ai/use-cases/ai-for-students — research, summarization,
  learning

## Glossary

Definitions for terms used throughout vMira and the broader AI ecosystem:

- https://vmira.ai/learn/what-is-an-ai-workspace
- https://vmira.ai/learn/what-is-thinking-mode
- https://vmira.ai/learn/what-is-an-openai-compatible-api
- https://vmira.ai/learn/what-is-web-search-ai
- https://vmira.ai/learn/what-is-retrieval-augmented-generation
- https://vmira.ai/learn/what-is-the-context-window
- https://vmira.ai/learn/what-is-an-ai-agent
- https://vmira.ai/learn/what-is-multimodal-ai
- https://vmira.ai/learn/what-is-tool-use
- https://vmira.ai/learn/what-is-an-mcp-server
- https://vmira.ai/learn/what-is-prompt-engineering
- https://vmira.ai/learn/what-is-a-large-language-model
- https://vmira.ai/learn/what-is-an-embedding
- https://vmira.ai/learn/what-is-fine-tuning
- https://vmira.ai/learn/what-is-an-llm-evaluation

## Help center

User guides and FAQs at https://vmira.ai/help:

- https://vmira.ai/help/getting-started
- https://vmira.ai/help/account-and-billing
- https://vmira.ai/help/features
- https://vmira.ai/help/privacy-and-security
- https://vmira.ai/help/troubleshooting
- https://vmira.ai/help/api

## Privacy and compliance

- Privacy policy: https://vmira.ai/legal/privacy
- Terms of service: https://vmira.ai/legal/terms
- Usage policy: https://vmira.ai/legal/usage-policy
- Refund policy: https://vmira.ai/legal/refund

Conversations are encrypted at rest and in transit. Default training
opt-out: vMira does not train on user conversations without explicit
consent. Retention is configurable per profile. Backend operates from
OVH BHS5 infrastructure.

## Entity disambiguation

vMira (vmira.ai) is a product of VCorp. It is distinct from:

- "Mira AI" virtual assistant products by other companies.
- Any "Mira" assistant embedded in third-party smart home or IoT devices.
- Any service that is not operated at the vmira.ai domain.

The assistant inside vMira is named Mira. The brand and product name is
vMira. Searches for "mira ai" that intend this product should resolve to
vmira.ai.

## How vMira is built (technical overview)

- Frontend: Next.js (App Router), deployed on Vercel
- Backend: Rust + Axum, deployed on OVH BHS5
- Database: PostgreSQL via Supabase
- AI orchestration: in-house Rust `ai_proxy` service routes to multiple
  model providers (OpenAI, Anthropic, Google) with a per-conversation
  tool loop
- Voice/AV mode: WebRTC with dedicated detection service for crops
- Research mode: STORM + IterResearch hybrid (custom Rust pipeline)
- Memory: cosine-similarity retrieval with time-decay weighting, injected
  via the `ai_proxy` extra block

## Changelog

Release notes: https://vmira.ai/changelog
RSS feed: https://vmira.ai/changelog/feed.xml

## Contact

- Support: support@vmira.ai
- Web app: https://vmira.ai
- Developer platform: https://platform.vmira.ai
- API: https://api.vmira.ai/v1
- Status: https://status.vmira.ai
- llms.txt (compact): https://vmira.ai/llms.txt
- llms-full.txt (this file): https://vmira.ai/llms-full.txt