VibePnL
Now with DeepSeek R1 + Ollama support

AI-powered
stock analysis.
Debate-tested.

Pick a ticker. Choose your AI provider. Our multi-agent system runs a bull-vs-bear debate and delivers a structured verdict — BUY, HOLD, or SELL — with full reasoning.

BUYOVERWEIGHTHOLDUNDERWEIGHTSELL
NVDANVIDIA Corp.
Live
Final Verdict4.2s · 18,421 tok
BUYConfidence: 87%
BULL

NVDA dominates AI infrastructure with 80%+ GPU market share. Data center revenue up 427% YoY.

BEAR

Valuation at 35x revenue is stretched. Competition from AMD, Intel intensifying.

viaclaude-sonnet-4-6
< 3 min
avg analysis time
5
AI providers
6+
specialized agents
100%
reasoning visible

How it works

From ticker to verdict in minutes

01
Pick a ticker

Enter any stock symbol — NVDA, AAPL, TSLA, SPY. Our system pulls current market context automatically.

02
Choose your AI

Select Claude, GPT-4, Gemini, DeepSeek, or a local Ollama model. Plug in your API key.

03
Get your verdict

Receive a structured BUY/HOLD/SELL rating with the full debate transcript, confidence scores, and token usage.

Features

Everything you need to analyze smarter

Multi-model support

Claude, GPT-4, Gemini, DeepSeek, Ollama — bring your own API key or use local models.

Multi-agent debate

Bull researcher vs bear researcher, risk managers, market analysts, and news analysts argue before a final verdict.

5 signal levels

BUY → OVERWEIGHT → HOLD → UNDERWEIGHT → SELL with confidence scores and nuanced reasoning.

Scheduled analysis

Set cron jobs to run analysis automatically. Get fresh insights on your watchlist daily or weekly.

Full reasoning trace

Every conclusion comes with the complete agent chain-of-thought, sources, and debate transcript.

Cost transparency

See exact token usage and cost per run. No surprises — you control the budget completely.

Multi-Agent System

6 specialized agents. One verdict.

Rather than a single LLM call, VibePnL orchestrates a structured debate between agents with opposing mandates.

Bull Researcher

Finds growth catalysts and bullish signals

Bear Researcher

Identifies risks, headwinds, and red flags

Risk Manager

Evaluates portfolio risk and position sizing

Market Analyst

Assesses macro conditions and sector trends

News Analyst

Processes recent news and sentiment signals

Synthesis Agent

Weighs all inputs and issues the final verdict

Signal confidence breakdown

BUY
87%
OVERWEIGHT
72%
HOLD
55%
UNDERWEIGHT
38%
SELL
20%
$ vibepnl analyze
ticker: NVDA
model: claude-sonnet-4-6
verdict: BUY
✓ 6 agents · 4.2s · 18,421 tokens · $0.042

AI Providers

Bring your own AI provider

VibePnL is model-agnostic. Use the frontier model you trust, or run everything locally with Ollama.

Anthropic
Claude
Ready
OpenAI
GPT-4
Ready
Google
Gemini
Ready
DeepSeek
DeepSeek R1
Ready
Ollama
Local Models
Ready

More providers coming soon. Open an issue or contribute on GitHub.

Built for serious analysis

Not a black box. VibePnL shows every step of the reasoning — you decide what to trust.

The bull-vs-bear debate format actually changed how I think about due diligence. It's not just a rating — it's a structured argument.

Alex M.
Retail investor

Being able to swap between Claude and GPT-4 on the same ticker and compare the reasoning is genuinely useful. The models disagree more than you'd think.

Sarah K.
Quant researcher

I set up a daily cron job for my 12-stock watchlist. Takes 4 minutes total and costs less than a coffee.

David R.
Portfolio manager

Research & Educational Use Only. VibePnL is a research and educational tool. Nothing on this platform constitutes financial advice. Always do your own research before making investment decisions.

Ready to analyze smarter?

No subscriptions. No black boxes. Just transparent AI-powered stock analysis.