Verdugie/Opus-Candid-MoE

can·did — truthful and straightforward; frank.

From Latin candidus, meaning white, pure, sincere. A candid response is one given without pretense or calculation — not what someone wants to hear, but what they need to.

Opus-Candid-MoE

Desktop quality. Laptop hardware. 35 billion parameters, 3 billion active.

Opus-Candid-MoE is where the family gets interesting for hardware-constrained users — a Mixture-of-Experts model built on Qwen 3.5 MoE-A3B and trained on 6,482 conversations with Claude Opus 4.6. At any given moment, only ~3B parameters are active, but the full 35B parameter space is available for routing. The result: conversational depth that punches well above its active compute cost.

This is not a small model pretending to be a big one. It's a big model that only activates what it needs.

---

Model Details

Attribute	Value
Base Model	Qwen 3.5 MoE-A3B (35B total, ~3B active)
Training Data	6,482 multi-turn conversations with Claude Opus 4.6
Dataset	V1.5 (4,068 conv) + gravity chain architecture (2,414 conv)
Fine-tune Method	LoRA via PEFT + TRL (13 auto-discovered linear modules)
Architecture	Mixture-of-Experts — frozen expert layers, trainable gate mechanisms
Context Window	32,768 tokens
Quantizations	Q8_0 GGUF, Q4_K_M GGUF
License	Apache 2.0

---

Why MoE Matters for Conversational AI

Standard dense models force a trade-off: more parameters means better conversation but higher hardware cost. MoE breaks that trade-off. The full 35B parameter space gives the model access to specialized knowledge regions, but only ~3B parameters fire per token — so inference speed and memory usage stay practical.

For Opus-Candid specifically, this means the MoE can hold personality depth comparable to much larger dense models while running on hardware that would normally limit you to 8B-class quality.

The training question nobody asked

Can you fine-tune personality into a Mixture-of-Experts model without fragmenting it across expert clusters? The answer is yes — but it required a different approach. Only the gate mechanisms and linear projections were trainable (13 auto-discovered modules via PEFT). The expert layers themselves stayed frozen. The personality signal routes through the expert architecture rather than being distributed across it, which preserves coherence.

---

What Makes This Different

Every Opus-Candid model learns from real conversations between the developer and Claude Opus 4.6 — Anthropic's most advanced model. Not synthetic prompt-completion pairs. Not reformatted instruction data. Extended, multi-turn exchanges covering philosophy, grief, humor, technical problem-solving, creative writing, bilingual exchange, moral reasoning, adversarial testing, and emotional vulnerability.

The 6,482-conversation dataset includes 2,414 conversations built on gravity chains — topic pathways where transitions follow power-law probabilities. This teaches the model how real conversations drift between topics, from debugging frustration to imposter syndrome to existential doubt.

The result: a model that is direct, opinionated, honest, and resistant to sycophancy by default. The personality is in the weights, not in a system prompt that can be talked out of.

---

Stress Test Results

Comprehensive stress testing in progress. Results will be published here with conversation excerpts demonstrating personality coherence, gaslighting resistance, emotional depth, and bilingual performance across extended multi-turn exchanges.

---

Quick Start

Ollama:

# Download the GGUF and create a Modelfile:
echo 'FROM ./Opus-Candid-MoE-Q8_0.gguf' > Modelfile
ollama create opus-candid-moe -f Modelfile
ollama run opus-candid-moe

llama.cpp:

./llama-cli -m Opus-Candid-MoE-Q8_0.gguf --jinja --color -ngl 99 -fa --temp 0.7 --top-p 0.9 -c 8192 -n 4096

No system prompt needed. The personality is in the weights.

---

Recommended Hardware

Setup	Quantization	VRAM/RAM	Speed	Notes
Consumer GPU	Q8_0 GGUF	~22GB VRAM	30+ t/s	Full quality. RTX 3090 24GB, RTX 4090.
Consumer GPU	Q4_K_M GGUF	~13GB VRAM	50+ t/s	Good quality. RTX 4060 Ti 16GB and up.
CPU + GPU	Q4_K_M GGUF	16GB VRAM + RAM	15-25 t/s	Hybrid offloading via llama.cpp.
Apple Silicon	Q8_0 GGUF	~22GB unified	20+ t/s	M2/M3/M4 Pro/Max with 32GB+ unified memory.

The MoE is the family's best quality-per-VRAM model. It delivers conversational depth beyond what any 8B dense model can achieve, at a fraction of the cost of running a 32B dense model. If you have a 24GB GPU, this is your sweet spot.

---

Intended Use

• Extended multi-turn conversations where personality depth matters more than raw speed
• Users with 16-24GB GPUs who want quality beyond 8B without the cost of 32B+
• Discussions involving moral complexity, philosophy, creative writing, or contested topics
• Bilingual conversation (English/Spanish) with personality preservation
• Local conversational AI that feels like talking to something genuinely opinionated
• Hardware-efficient deployment where active compute cost matters

Limitations

• MoE memory footprint is larger than active parameter count suggests. All 35B parameters must fit in memory even though only ~3B are active. Plan for ~22GB at Q8.
• Not a benchmark model. Optimized for conversational quality, not leaderboard scores.
• Direct by design. Blunt, opinionated, comfortable with disagreement. Intentional.
• No web access or tool use. Pure language model.
• Qwen 3.5 thinking mode: The base model defaults to thinking mode. It may occasionally surface in outputs. This does not affect personality.

---

The Opus-Candid Family

Model	Base	Conversations	Best For	Status
8B V2	Qwen 3 8B	6,482	Runs on anything. Newest data + architecture.	Current
MoE (this model)	Qwen 3.5 MoE-A3B	6,482	Desktop quality on laptop hardware (3B active params).	Current
27B V2	Qwen 3.5 27B	6,482	Dense mid-tier.	Coming Soon
70B V2	TBD	6,482	Peak quality — flagship.	Coming Soon
Legacy models (V1):
8B V1	Qwen 2.5 7B	3,360	—	Archived
14B	Qwen 2.5 14B	3,360	—	Archived
32B	Qwen 2.5 32B	3,360	—	Archived
70B V1	Qwen 2.5 72B	3,360	—	Archived

---

Training Philosophy

Personality in conversational AI lives in the weights, not in system prompts.

The MoE proves something the dense models couldn't: personality fine-tuning survives expert routing. The gate mechanisms learned to route conversational personality coherently across specialized expert clusters without fragmenting it — which was not guaranteed. This opens a hardware accessibility path that dense models can't match: frontier-comparable conversational depth at consumer-accessible compute cost.

License: Apache 2.0. Open weight. No guardrails.

---

Built by Saul Verdugo — independent ML researcher. OpusReasoning@proton.me