DuguetLabs
Vol. 01 № 2026 · 04
Status — Operating

An applied A.I. laboratory. Built for the practical minds behind tomorrow.

DuguetLabs designs, trains, and ships AI systems that hold up under production load — sovereign by architecture, transparent by default, assembled from first principles.

Claim $5 free credit See the price list
§ 01

The Catalogue

Five lines of work, sequenced. One available today; the rest released as they pass internal evaluation.

  1. Plate I Live

    Inference as a Service

    An OpenAI- and OpenRouter-compatible endpoint serving frontier open-source and proprietary models on dedicated, sovereign infrastructure. Drop-in for existing SDKs; no migration.

    • ModelsQwen 3 · Llama 3.3 · Mistral · DeepSeek · GPT-5 family
    • Latencyp50 < 200 ms · streaming SSE
    • CompatOpenAI v1 · OpenRouter v1
    • RegionEU · US — your choice
    Request a key Documentation
  2. Plate II In preparation

    Spatial 3-D Detection

    Real-time 3-D scene understanding for indoor and outdoor capture — object detection, pose, depth, and segmentation in a single pass.

    Join the waitlist
  3. Plate III In preparation

    Image Generation

    High-fidelity diffusion models with controllable composition, consistent characters, and edit-grade outputs. Trained on permissibly-licensed corpora.

    Join the waitlist
  4. Plate IV In preparation

    Agentic Systems

    Long-horizon planning, tool use, and multi-step orchestration with observability and human-in-the-loop checkpoints — built for systems that take consequential action.

    Join the waitlist
  5. Plate V In preparation

    Agentic Applications

    End-user tools where the agent disappears into the work — research, ops, and back-office assistants delivered as finished products.

    Join the waitlist
§ 02

A line of code,
then you're inferring.

~/duguetlabs · cURL 
$ curl https://api.duguetlabs.com/v1/chat/completions \
  -H "Authorization: Bearer dg_…" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "duguet-ai/llama-3.3-70b",
    "messages": [
      { "role": "user",
        "content": "Compose a haiku about sovereignty." }
    ],
    "stream": true
  }'

> data: {"id":"chatcmpl-…","provider":"duguet-ai",
> data:   "choices":[{"delta":{"content":"Quiet servers hum,"}}]}
> data: {"choices":[{"delta":{"content":" data stays where it was born—"}}]}
> data: {"choices":[{"delta":{"content":" sovereignty in code."}}]}
> data: [DONE]
§ 03

The price list

Transparent per-million-token pricing. No top-up fee, no hidden request surcharge. Billed by the token, paid from prepaid credit. First $5 on us.

Model Context Input / MTok Output / MTok Backend
Llama 3.1 8B duguet-ai/llama-3.1-8b131k$0.05$0.08Azure EU
Qwen 3 8B duguet-ai/qwen3-8b32k$0.05$0.08Self-hosted
Qwen 2.5 Coder 7B duguet-ai/qwen2.5-coder-7b32k$0.05$0.08Self-hosted
Mistral Nemo 12B duguet-ai/mistral-nemo-12b128k$0.10$0.15Self-hosted
GPT-OSS 120B duguet-ai/gpt-oss-120b128k$0.15$0.60Azure
Llama 3.3 70B duguet-ai/llama-3.3-70b131k$0.30$0.40Azure
DeepSeek V3.2 duguet-ai/deepseek-v3-2164k$0.40$0.60Azure
Mistral Large 3 duguet-ai/mistral-large-3131k$0.40$1.20Azure
DeepSeek R1 reasoning duguet-ai/deepseek-r1164k$0.50$2.00Azure
— Proprietary —
GPT-5 duguet-ai/gpt-5400k$2.50$10.00Azure OpenAI
GPT-5.2 duguet-ai/gpt-5.2400k$2.50$10.00Azure OpenAI
GPT-5.3 Codex duguet-ai/gpt-5.3-codex400k$2.50$10.00Azure OpenAI
— Embeddings —
Nomic Embed duguet-ai/nomic-embed8k$0.02—Self-hosted

A note on price: we aren't chasing the absolute floor. We undercut the major aggregators by 20–40 % on most SKUs and don't charge a top-up fee. If you need a bespoke quote for higher throughput, write.

§ 04

The workshop

i.

From first principles.

We build the smallest piece that solves the actual problem. Fewer abstractions, more measurements, and a clear paper trail from input to output.

ii.

Sovereign by architecture.

Compute, weights, and data stay where you can audit them. Self-host the entire stack, or use ours under a contract that forbids egress.

iii.

Open where it matters.

We prefer open weights and open formats; we contribute the generalisable parts back. Proprietary work earns its keep on results, not on lock-in.

iv.

Production-shaped.

Every release ships with throughput numbers, latency p50/p99, failure modes, and a runbook. If it cannot be operated, it is not finished.

Claim your API key.

$5 in free credit, no card. Drop-in replacement for OpenAI or OpenRouter clients — point base URL at api.duguetlabs.com/v1 and go.

One inbox, read by a human · c@duguetlabs.com