Stop Choosing Models.
Start Building.
MisterPilot is the AI coding assistant for VS Code that picks the right model for every request — GPT, Claude, or DeepSeek — automatically. Fast answers on simple tasks, deep reasoning on hard ones, and you only pay for what you use.
- No subscription
- Bring your own keys
- Up to 1M-token context
- Copilot Chat compatible
›Why is this regex not matching?
Simple question
›Add Google OAuth to the Express app
Multi-file feature
›Split this monolith into services
Architecture task
Built for the tools developers already use
VS Code
Native extension
Copilot Chat
Compatible
OpenAI
GPT models
Claude
Anthropic models
DeepSeek
Chat & Reasoner
MCP
Model Context Protocol
AI That Picks The Right AI
Every model has strengths. Some are fast and cheap. Some reason deeply. Some handle huge architectural changes. Instead of making you guess, Auto reads each request and sends it to the model best suited for the job.
01
You ask
Chat, select code, or run a command — exactly like you do today.
02
Auto analyzes
MisterPilot looks at the task's complexity, context size, and type of work.
03
The right model answers
Your request goes to the best-fit model, and you can see which one answered.
“What does this function return?”
Simple question — speed matters more than depth.
“Implement pagination for this API”
Multi-file change with logic that has to be right.
“Plan a migration from REST to GraphQL”
Architectural decision — quality matters most.
Prefer a specific model? Pick it manually any time. Auto is the default, never a lock.
Stop Paying Frontier Prices
For Simple Questions
Most coding requests don't need the most expensive model. Auto uses premium models only when they make a difference.
Better answers
Hard problems get the strongest reasoning models — not whatever you last had selected.
Lower costs
Everyday questions run on fast, affordable models. Your wallet goes further.
No model-switching
No dropdowns, no second-guessing, no “should I have used Claude for this?”
Fully automatic
Works out of the box. Override any time if you want a specific model.
| Picking models yourself | MisterPilot Auto | |
|---|---|---|
| Model choice per request | You guess | Chosen from the task |
| Simple questions | Often sent to expensive models | Routed to fast, low-cost models |
| Complex tasks | Often sent to weaker models | Routed to reasoning & frontier models |
| Keeping up with new models | You research every release | Routing updated as models ship |
| Cost visibility | Separate provider dashboards | Every request in one dashboard |
All Your AI Models. One Workspace.
OpenAI, Claude, and DeepSeek in one extension, one chat, one bill. No vendor lock-in — when a better model ships, you get it.
OpenAI
Versatile, fast GPT models for everyday coding and quick edits.
Claude
Careful reasoning and large-context code understanding from Anthropic.
DeepSeek
Strong coding performance at a fraction of the cost.
Use Your Own AI Keys.
Already paying OpenAI, Anthropic, or DeepSeek? Connect your API keys and MisterPilot uses them directly — billed by the provider, at their rates. Or skip keys entirely and use MisterPilot credits.
Full control
Your keys, your provider accounts, your limits.
Direct billing
Usage goes straight to your provider invoice.
Use what you have
Existing credits and subscriptions just work.
Stored securely
Keys encrypted and managed with AWS Secrets Manager.
OpenAI
sk-••••••••••4f2a
Anthropic
sk-ant-••••••••91c
DeepSeek
MisterPilot credits
The AI Coding Assistant
That Lives in Your Editor
Install one extension. Keep your editor, your keybindings, and your workflow. MisterPilot works with the Copilot Chat experience you already know.
Ask anything about your code. Answers stream in real time.
verifyToken()decodes the JWT and comparesexpwithDate.now() / 1000.- Expired tokens throw
TokenExpiredError, caught inauthGuard→ 401.
Already use Copilot Chat?
MisterPilot plugs into the same chat experience — no new UI to learn. Just more models and smarter routing.
Your AI understands your workspace
Less prompting, more accurate answers — context is picked up automatically.
More Than Chat.
MisterPilot doesn't just answer — it reads your project, uses tools, and works through multi-step tasks.
MCP
Connect AI To Everything
With the Model Context Protocol, MisterPilot can talk to your databases, APIs, browsers, and internal services. Add a server — your AI gains a new skill.
Local agent execution
Your Agent Runs Where Your Code Lives
The agent runs inside VS Code, right next to your files — not on a remote backend for every action. Only model calls leave your machine.
Typical remote agent
MisterPilot
- ✓ Faster responses
- ✓ Better reliability
- ✓ More control
Up to 1,000,000 tokens
Built For Large Codebases
Huge codebases and long conversations stay in view — better project awareness, less repeating yourself, and long sessions that stay coherent.
8K
A file
128K
A module
200K
A service
1M
Your whole repo
Available context depends on the model in use.
Diagnostics awareness
Sees your errors and warnings, fixes what's actually broken.
File discovery
Finds the right files without you pointing at them.
Directory browsing
Understands how your project is organized.
Large file reading
Handles big files without losing what matters.
Multi-step tool use
Chains searches, reads, and edits to finish real tasks.
Automation Without Losing Control
Always AllowAuto-approve safe operations — like reading files — while anything risky still asks first. Faster workflows, safety by default.
- ✓ Faster workflows
- ✓ Safety maintained
Built For Production.
The boring parts, done right — so your AI assistant is there when you need it, and you always know what it costs.
Automatic fallbacks
If a provider has trouble, requests reroute to a healthy model.
Real-time streaming
Responses appear as they're generated. No waiting for the full answer.
Usage tracking
Every request logged with model, tokens, and cost.
Reliable billing
Prepaid wallet, accurate per-token charges, full transaction history.
Secure secrets
API keys encrypted and stored in AWS Secrets Manager.
Secure sign-in
Google or email auth, managed API keys with instant revoke.
Wallet Balance
₹1,250.00
API Requests
12,847
Tokens Used
2.4M
Amount Spent
₹124.50
Your usage dashboard: which model handled each request and exactly what it cost.
curl https://engine.misterpilot.online/api/v1/chat/completions \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek-chat",
"messages": [
{"role": "user", "content": "Hello!"}
]
}'Built For Developers
If you've used the OpenAI SDK before, you already know how to use MisterPilot. Change the base URL, add your key, pick a DeepSeek model — done.
OpenAI-Compatible API
Works with any OpenAI SDK — just point it at MisterPilot and go.
Fast Integration
Make your first API call in under 2 minutes.
Simple Authentication
One bearer token for all DeepSeek models.
Detailed Usage Tracking
Every token logged, every cost visible in your dashboard.
Pay for Tokens, Not Subscriptions
No complicated tiers. No surprise bills. Recharge your wallet or bring your own keys — and Auto keeps every rupee working harder.
MisterPilot Credits
Most popularMinimum recharge ₹99. Pay as you go — only for actual usage.
- No monthly subscription
- No hidden charges — recharge when needed
- OpenAI, Claude & DeepSeek models
- MisterPilot Auto routing
- MCP, agent tools & up to 1M context
- Real-time usage & cost dashboard
- API key management
No subscription. Recharge only when you need to.
Bring Your Own Key
Full controlAlready have provider accounts? Use them directly.
- Use your OpenAI, Anthropic & DeepSeek keys
- Billed directly by your provider
- Keys encrypted in AWS Secrets Manager
- Same extension, Auto routing & tools
- Mix: your keys + MisterPilot credits
MisterPilot vs typical subscription AI coding tools
| MisterPilot | Subscription tools | |
|---|---|---|
| Monthly subscription | None — from ₹99, pay as you go | Usually required |
| Pay only for what you use | ||
| OpenAI + Claude + DeepSeek in one place | Varies | |
| Automatic model routing | Varies | |
| Bring your own API keys | Varies | |
| Works in your existing VS Code | Varies | |
| Per-request cost visibility |
Frequently Asked Questions
Auto is our automatic model router. For each request it looks at what you're asking and sends it to the best-fit model — a fast, low-cost model for simple questions, a strong reasoning model for implementation work, and a frontier model for big architectural tasks.
Your Next Commit Deserves
The Right Model
Install MisterPilot in under a minute. No subscription, no model-juggling — just better code, faster.
Start with ₹99. Pay as you go.