openscribe/site/pages/connect-your-ai.html
Laurence 8fd9b694f8
Some checks failed
ci / firmware (pull_request) Failing after 31s
ci / openapi (pull_request) Failing after 33s
ci / emulator (pull_request) Failing after 24s
ci / server (pull_request) Failing after 37s
docs+site: Nightjar brand, hosted service, infrastructure plan, saved site
Persists the branding/commercial/infrastructure work (previously only in chat and
on the live site) into version control.

What changed:
- docs/naming.md: due-diligence and the decision to brand the product Nightjar
  (OpenScribe domain/GitHub/trademark collisions; alternatives screened; why Nightjar).
- docs/hosted-service.md: the commercial model - plan tiers (Self-host / Cloud
  Starter / Cloud Pro / Private), indicative pricing, backend mapping, and the
  build-to-sell roadmap (metering, billing, provisioning, DPA).
- docs/infrastructure.md: backend infra plan. Primary option = self-host on the
  3x Minisforum MS-02 cluster (one with an RTX 3050 6GB) fronted by a Cloudflare
  Tunnel, with node roles, the 3050 capacity reality, caveats, and a Hetzner cloud
  fallback. Provision-later.
- site/: reproducible marketing-site source - block content for all 11 pages
  (rebranded to Nightjar), the navigation, the Contact Form 7 config (honeypot),
  the ApisCP SOAP helper (tools/apiscp.php, no secret), and a README on how the
  WordPress site is built and managed via the API.
- state/: DECISIONS (Nightjar rebrand, hosted service, MS-02 backend), PROJECT
  (brand + commercial section), TODO (rename decision, trademark, mailbox, pricing,
  hosted-service build, infra provisioning).

Why:
- User asked to save everything to the repo. Captures the product rebrand, the
  commercialisation plan, and the infrastructure decision so a cold session has the
  full picture.

Notes:
- The repo is still named `openscribe`; the product/brand is Nightjar. A full
  codebase rename is deferred (tracked in TODO + docs/naming.md).
- No secrets committed: the ApisCP API key is read from a local scratch file, never
  the repo.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-05 22:24:24 +01:00

24 lines
4.7 KiB
HTML

<!-- wp:heading {"level":1} --><h1 class="wp-block-heading">Connect your AI</h1><!-- /wp:heading --><!-- wp:paragraph {"fontSize":"large"} --><p class="has-large-font-size">Nightjar does not hard-wire a single AI vendor. Transcription and summaries each target a provider you choose in the server config. This guide shows the exact settings for each.</p><!-- /wp:paragraph -->
<!-- wp:paragraph --><p>All settings live in the server&#8217;s <code>.env</code> file. Change them, restart the server, and the new provider is live. Nothing else needs to change on the device or in the app.</p><!-- /wp:paragraph -->
<!-- wp:heading --><h2 class="wp-block-heading">Summaries and chat (the language model)</h2><!-- /wp:heading --><!-- wp:paragraph --><p>Set <code>LLM_PROVIDER</code> to one of <code>ollama</code>, <code>openai_compatible</code>, or <code>anthropic</code>.</p><!-- /wp:paragraph -->
<!-- wp:heading {"level":3} --><h3 class="wp-block-heading">Fully local (Ollama)</h3><!-- /wp:heading --><!-- wp:paragraph --><p>The default. Everything stays on your machine, no API key, no cloud. Install <a href="https://ollama.com">Ollama</a>, pull a model, and point Nightjar at it.</p><!-- /wp:paragraph --><!-- wp:code --><pre class="wp-block-code"><code>NIGHTJAR_LLM_PROVIDER=ollama
NIGHTJAR_OLLAMA_URL=http://localhost:11434
NIGHTJAR_OLLAMA_MODEL=llama3.1</code></pre><!-- /wp:code -->
<!-- wp:heading {"level":3} --><h3 class="wp-block-heading">OpenAI, or anything OpenAI-compatible</h3><!-- /wp:heading --><!-- wp:paragraph --><p>One setting covers OpenAI, Groq, OpenRouter, Together, LocalAI, vLLM, LM Studio, and Gemini&#8217;s OpenAI-compatible endpoint. Just change the base URL and model.</p><!-- /wp:paragraph --><!-- wp:code --><pre class="wp-block-code"><code>NIGHTJAR_LLM_PROVIDER=openai_compatible
NIGHTJAR_LLM_BASE_URL=https://api.openai.com/v1
NIGHTJAR_LLM_API_KEY=sk-...
NIGHTJAR_LLM_MODEL=gpt-4o-mini</code></pre><!-- /wp:code --><!-- wp:paragraph --><p>For Groq (fast and cheap), use <code>https://api.groq.com/openai/v1</code>. For a local model server on your LAN such as LM Studio, use <code>http://localhost:1234/v1</code>.</p><!-- /wp:paragraph -->
<!-- wp:heading {"level":3} --><h3 class="wp-block-heading">Anthropic Claude</h3><!-- /wp:heading --><!-- wp:paragraph --><p>Claude uses its own integration.</p><!-- /wp:paragraph --><!-- wp:code --><pre class="wp-block-code"><code>NIGHTJAR_LLM_PROVIDER=anthropic
NIGHTJAR_LLM_API_KEY=sk-ant-...
NIGHTJAR_LLM_MODEL=claude-opus-4-8</code></pre><!-- /wp:code -->
<!-- wp:heading --><h2 class="wp-block-heading">Transcription (speech to text)</h2><!-- /wp:heading --><!-- wp:paragraph --><p>Set <code>TRANSCRIPTION_PROVIDER</code> to <code>local_whisper</code> (self-hosted) or <code>openai_compatible</code>.</p><!-- /wp:paragraph -->
<!-- wp:heading {"level":3} --><h3 class="wp-block-heading">Local, self-hosted (faster-whisper)</h3><!-- /wp:heading --><!-- wp:code --><pre class="wp-block-code"><code>NIGHTJAR_TRANSCRIPTION_PROVIDER=local_whisper
NIGHTJAR_WHISPER_MODEL=base
NIGHTJAR_WHISPER_DEVICE=cpu
NIGHTJAR_WHISPER_COMPUTE_TYPE=int8</code></pre><!-- /wp:code --><!-- wp:paragraph --><p>Larger models (<code>small</code>, <code>medium</code>, <code>large-v3</code>) are more accurate but slower. Set <code>DEVICE=cuda</code> if you have a GPU.</p><!-- /wp:paragraph -->
<!-- wp:heading {"level":3} --><h3 class="wp-block-heading">A hosted transcription API</h3><!-- /wp:heading --><!-- wp:code --><pre class="wp-block-code"><code>NIGHTJAR_TRANSCRIPTION_PROVIDER=openai_compatible
NIGHTJAR_TRANSCRIPTION_BASE_URL=https://api.openai.com/v1
NIGHTJAR_TRANSCRIPTION_API_KEY=sk-...
NIGHTJAR_TRANSCRIPTION_MODEL=whisper-1</code></pre><!-- /wp:code -->
<!-- wp:heading --><h2 class="wp-block-heading">Which should I choose?</h2><!-- /wp:heading --><!-- wp:list --><ul class="wp-block-list"><!-- wp:list-item --><li><strong>Maximum privacy, zero cost per minute:</strong> local Whisper + a local Ollama model. Nothing leaves your machine.</li><!-- /wp:list-item --><!-- wp:list-item --><li><strong>Best quality, least setup:</strong> a commercial API for both. You pay per use, and accuracy is top-tier.</li><!-- /wp:list-item --><!-- wp:list-item --><li><strong>A balance:</strong> local transcription for privacy, a commercial model for the summary polish.</li><!-- /wp:list-item --></ul><!-- /wp:list -->
<!-- wp:paragraph --><p>Not sure, or would rather not manage any of this? <a href="/hosted-plans/">Our hosted plans</a> run it all for you.</p><!-- /wp:paragraph --><!-- wp:buttons --><div class="wp-block-buttons"><!-- wp:button --><div class="wp-block-button"><a class="wp-block-button__link wp-element-button" href="/self-host-the-server/">Next: self-host the server</a></div><!-- /wp:button --></div><!-- /wp:buttons -->