Multi-model deliberation CLI. 5 frontier LLMs debate your question, then Claude Opus 4.6 judges and synthesizes a recommendation.
- Blind phase — Each model answers independently (no herding)
- Cross-pollination — Models read all blind claims and investigate gaps
- Debate — Structured rounds with a rotating challenger ensuring sustained disagreement
- Judge — Claude Opus synthesizes using Analysis of Competing Hypotheses
- CollabEval — Gemini critiques the judge's synthesis; judge revises
- Extraction — Structured Do Now / Consider Later / Skip recommendations
Auto-routes by difficulty: simple questions get quick parallel queries ($0.10), complex ones get full council deliberation ($0.50).
Multi-model collaboration produces better outcomes than any single model — but naive "ask multiple models and vote" fails. Most of the benefit from debate comes from majority voting alone; the iterative discussion is a martingale that doesn't improve expected correctness without targeted interventions. consilium implements those interventions:
| Design principle | What it prevents | Evidence |
|---|---|---|
| Cross-lab panel (5 labs) | Shared training misconceptions reinforcing wrong answers | ReConcile, ACL 2024 — +11.4% over homogeneous MAD |
| Blind first phase | Identity/authority anchoring (Compassion-Fade bias) | CALM, 2024 |
| Anonymous labels (Speaker 1/2/3) | Judge favoring named model outputs | IBC paper, 2024 — conformity > obstinacy when identity known |
| Rotating challenger | Sycophancy cascades; models converging to majority | DEBATE, ACL 2024 |
| Anti-capitulation prompt | Position changes without new evidence | Peacemaker/Troublemaker, 2024 — r=0.902 sycophancy↔wrong answer |
| 1–2 round cap | Sycophancy intensifies at round 3+ | Peacemaker/Troublemaker, 2024 |
| Judge outside panelist pool | Self-Enhancement (judge favoring same-lab outputs) | CALM, 2024 |
| Confidence extraction for judge | Equal-weighting confident and uncertain positions | ReConcile, ACL 2024 |
| Gemini critique + judge revision | Fallacy-Oversight (worst judge bias, 0.566 score) | CALM, 2024 |
| Role | Model | Lab |
|---|---|---|
| Panelist | GPT-5.2 Pro | OpenAI |
| Panelist | Claude Sonnet 4.6 | Anthropic |
| Panelist | Grok-4.20 | xAI |
| Panelist | DeepSeek V3.2 | DeepSeek |
| Panelist | GLM-5 | Zhipu |
| Judge | Gemini 3.1 Pro | |
| Critique | Claude Opus 4.6 | Anthropic |
5 distinct labs on the panel. The judge (Google) is not in the panelist pool.
# Build from source
cargo build --release
# Binary at target/release/consilium
# Optionally symlink:
ln -s $(pwd)/target/release/consilium ~/.local/bin/consiliumRequires OpenRouter API key:
export OPENROUTER_API_KEY=sk-or-v1-...
export GOOGLE_API_KEY=AIza... # optional, Gemini fallback# Auto-route (Opus picks the best mode)
consilium "Should I take this job offer?"
# Quick parallel — independent opinions, no debate (~$0.10)
consilium "Best practices for error handling in Rust?" --quick
# Full council with JSON output (~$0.50)
consilium "Microservices vs monolith for a 5-person startup?" --council --format json
# Deep — auto-decompose + 2 debate rounds (~$0.90)
consilium "How should banks govern AI agents?" --deep
# Oxford debate — binary for/against + verdict (~$0.40)
consilium "Should we adopt Kubernetes?" --oxford
# Red team — adversarial stress-test (~$0.20)
consilium "My plan: rewrite the backend in Rust over 6 months" --redteam
# Roundtable discussion (~$0.30)
consilium "Future of open-source AI models?" --discuss --rounds 2
# Socratic examination (~$0.30)
consilium "Is consciousness computable?" --socratic| Mode | Flag | Cost | Description |
|---|---|---|---|
| Auto | (default) | varies | Opus classifies and picks the best mode |
| Quick | --quick |
~$0.10 | Parallel queries, no debate |
| Council | --council |
~$0.50 | Full multi-round deliberation + judge |
| Deep | --deep |
~$0.90 | Council + decompose + 2 rounds |
| Oxford | --oxford |
~$0.40 | Binary for/against debate |
| Red Team | --redteam |
~$0.20 | Adversarial stress-test |
| Discuss | --discuss |
~$0.30 | Hosted roundtable |
| Socratic | --socratic |
~$0.30 | Assumption-probing examination |
--persona "context" Personal context injected into prompts
--domain banking|healthcare Domain-specific regulatory context
--challenger gemini Assign contrarian role (council mode)
--decompose Break question into sub-questions first
--xpol Cross-pollination phase (council mode)
--followup Interactive drill-down after synthesis (council mode)
--rounds N Rounds for discuss/socratic (0 = unlimited)
--effort low|medium|high Per-phase reasoning effort for thinking models
--thorough Disable early consensus exit
--output file.md Save transcript to file
--format json|yaml|prose Output format
--share Upload to secret GitHub gist
--quiet Suppress live output (auto-enabled when not a TTY)
--no-save Don't auto-save session
consilium --stats # Cost breakdown by mode
consilium --sessions # List recent sessions
consilium --view # View latest in pager
consilium --view "career" # View session matching term
consilium --search "AI" # Search session content
consilium --watch # Live tail (styled terminal)
consilium --tui # Full TUI viewer
consilium --doctor # Check API keys and connectivity~6,100 lines of Rust. Single 4.7MB binary, ~50ms cold start.
- Single tokio runtime, async throughout
- SSE streaming with
<think>block filtering (DeepSeek-R1, OpenAI reasoning) - CostTracker via AtomicU64 (micro-dollars, lock-free across tasks)
- Output trait enables TeeOutput (stdout + live file) for watch/TUI
- LiveWriter with PID-based file management and stale cleanup
Rewritten from consilium-py (Python, 5,091 lines). Same CLI interface, same modes, same output format, same ~/.consilium/ session directory. Both versions can coexist and read each other's sessions.
MIT