/Live_Demo — Read-only

// Council_Mode

One prompt. Three frontier models answer in parallel. A judge model synthesizes a single consensus answer, scores agreement, and maps every point of dissent. This is a real-shaped sample output — on your account it runs against your own API keys in ~10 seconds.

Prompt_
You are a senior investment analyst. A 38-year-old software engineer in Bangalore has ₹40L in cash savings and a stable ₹3L/month income. Recommend a 5-year asset allocation balancing long-term growth with downside protection. Justify your picks.
//

Synthesis_

judge_verdict
84%Agreement
Consensus_

All three models converge on a 60/25/10/5 equities / debt / alternatives / cash starting allocation, with slow glide-path rebalancing toward 50/35/10/5 by year five.

Equities emphasis:

  • Domestic large-cap via low-cost index funds (Nifty 50 or Nifty Next 50) as the core holding — 35% of portfolio.
  • International equity (S&P 500 via FoF or direct US brokerage) for currency + geography diversification — 15%.
  • Flexi-cap or mid-cap active fund for a measured growth tilt — 10%.

Debt and protection:

  • PPF + EPF + debt MFs laddered across 3-, 5-, and 7-year durations. Models disagree on whether to prefer corporate bond funds or gilt funds inside this sleeve.
  • 6-month emergency fund held in a liquid fund, ring-fenced from the allocation.

Alternatives:

  • REITs (Embassy/Mindspace) for rental-like cash flow without concentrated real estate risk.
  • Gold ETF or SGB at 5% as a macro hedge.

All models recommend quarterly review, annual rebalance, and avoiding direct equity stock picking. They flag tax-harvesting as a meaningful alpha lever at this capital level.

/Dissent_Map
Crypto / BTC exposure
gpt-5.2Allocate a satellite 2–3% to BTC via a regulated ETF for asymmetric upside; cap it and rebalance strictly.
claude-opus-4-5Avoid crypto entirely at this life stage — drawdown risk and custody complexity outweigh expected return for a family with concentrated household income.
gemini-3-pro-previewOptional 1% exposure only after the emergency fund, insurance, and PPF ceilings are fully utilized.
Debt fund flavor
gpt-5.2Prefer short- and medium-duration corporate bond funds for carry; gilts only as a tactical duration bet.
claude-opus-4-5Weight gilt funds higher for credit safety; accept lower yield in exchange for sovereign backing.

// Member_Responses

3 models
OpenAIgpt-5.2
stop
Starting allocation: 60% equities / 25% debt / 10% alternatives / 5% cash.

Equities: 35% domestic large-cap index (Nifty 50 via UTI/HDFC index fund), 15% S&P 500 via FoF, 10% flexi-cap active. Debt: 15% short-duration corporate bonds, 10% PPF/EPF topped up to ₹1.5L/yr ceilings. Alternatives: 5% REITs (Embassy, Mindspace), 5% gold (SGB > ETF because of tax treatment at maturity). Cash: 6-month liquid fund emergency reserve.

Consider a 2–3% satellite BTC allocation via a regulated ETF once the a
Latency
8.90s
Tokens
2,780
Cost
$0.034
Anthropicclaude-opus-4-5
stop
I'd anchor this around a 60/25/10/5 split. The engineer's income stability justifies a growth tilt, but Indian household concentration risk argues for geographic diversification.

Recommend: 35% Nifty 50 index (direct, low ER), 15% international equity via a US-listed S&P 500 ETF (ideally via LRS rather than an Indian FoF to avoid the debt-fund taxation trap since 2023). 10% mid-cap active. Debt side: 20% PPF + EPF + gilt funds (I prefer gilts here — sovereign safety matters more than yield when
Latency
11.20s
Tokens
3,310
Cost
$0.052
Googlegemini-3-pro-preview
stop
Allocate 60/25/10/5. The key insight for Indian investors at this capital level is tax-efficient structure, not stock picking.

Equity sleeve: ELSS (until ₹1.5L 80C is exhausted), Nifty 50 index, and a Nasdaq-100 ETF for tech exposure since the user already has human capital in the same sector (consider underweighting tech for correlation reasons). Debt: combination of PPF ceiling, corporate bond fund, and a 5-year FD ladder for predictable cash flow. 5% REITs + 5% gold. Cash at 5% in a sweep ac
Latency
7.60s
Tokens
2,330
Cost
$0.0080
Total_Cost
$0.094
Total_Tokens
8,420
Latency
11.20s
Models
3