Providers & models
Briglia's main agent runs on the provider you choose; a few side services (voice, OCR, search, page reading) have their own keys. You can configure several main providers and switch with /provider at any time.
Main agent providers
OpenCode Go (recommended)
One subscription key, generous limits, a curated catalogue where every model has been tested with Briglia's tools and reasoning replay. Endpoint https://opencode.ai/zen/go/v1.
| Model id | Name | Notes |
|---|---|---|
glm-5.3-flash | GLM 5.3 Flash | vision · default |
kimi-k2.6 | Kimi K2.6 | vision |
glm-5.3 | GLM 5.3 | text-only |
deepseek-v4-pro | DeepSeek V4 Pro | text-only |
deepseek-v4-flash | DeepSeek V4 Flash (requires China opt-in) | text-only |
deepseek-v4-flash-vision-exp | DeepSeek V4 Flash Vision (experimental) | vision |
gpt-5.6-luna | GPT 5.6 Luna | text-only |
kimi-k2.7-code | Kimi K2.7 Code | vision |
kimi-k3 | Kimi K3 | vision |
minimax-m3 | MiniMax M3 | vision |
qwen3.8-max | Qwen 3.8 Max | vision |
OpenRouter
Pay-per-token access to almost any model; enter any model id (the wizard suggests google/gemini-3-flash-preview). The wizard asks whether the model can see images.
Custom OpenAI-compatible endpoint
Any server speaking the OpenAI chat-completions API: base URL, model id and API key. Reasoning models are supported, with per-provider replay of their reasoning across turns.
Local server
vLLM, Ollama, LM Studio and friends on your own hardware — no key, default http://localhost:1234/v1. Prefix caching on the server helps a lot with Briglia's long contexts.
With a text-only main model, images and scanned PDFs go through an OCR preprocessor first (on your OpenAI key), so they still work — a vision model just sees them directly. PDFs are rendered page by page at a bounded resolution and cached by content hash, so a 20-page document doesn't become tens of megabytes on every request.
Side services
| Service | Used for | Where to get it |
|---|---|---|
| OpenAI | web-research planning and writing, voice transcription (gpt-transcribe), image generation (gpt-image-2), OCR of scanned documents | platform.openai.com |
| Serper | Google search results | serper.dev (2,500 free queries) |
| Jina Reader | clean text of web pages | jina.ai/reader (free tier) |
The web-research backend (which model plans and synthesises research) can be OpenAI, OpenRouter or OpenCode — /websearch shows or switches it.
Subagent lanes
Briglia can delegate to subagents (exploration, planning, browsing, watcher triage). Each runs on a lane: inherit (the main model, default), cheap-vision or cheap-text. Configure the two cheap lanes once with /subagentmodels per provider; unset lanes fail loudly rather than silently falling back. /subagents off removes delegation entirely.
Costs
Briglia has no subscription of its own — you pay your providers. It is built to be frugal: exact token accounting from provider telemetry, prompt-cache-friendly request shapes, compaction instead of blind truncation, and optional per-turn, daily and monthly spend limits you set from chat (/spend) — none are on by default.