Returns an ellmer chat object configured for the given
provider, ready to be passed to the extraction functions
(extract_horizons_from_pdf, etc.). The chat object
wraps API credentials and model selection; it does not itself send
any request.
Usage
vlm_provider(
name = c("auto", "anthropic", "openai", "google", "ollama"),
model = NULL,
...
)Arguments
- name
Provider name. One of
"anthropic"(Claude),"openai"(GPT-4o family),"google"(Gemini),"ollama"(local).- model
Optional model identifier; defaults to
default_model(name).- ...
Additional arguments forwarded to the corresponding
ellmer::chat_*constructor (e.g.system_prompt,api_key,base_url,params).
Details
This is purely a convenience wrapper: it picks a default model per
provider and forwards remaining arguments (e.g.
system_prompt, api_key) to the underlying ellmer
constructor. ellmer must be installed.
Local-first option
Passing name = "ollama" runs every extraction locally via
an Ollama server (default gemma4:e4b, Gemma 4 edge with
multimodal text+image+audio support). No data leaves the
machine, which is the recommended setting for sensitive field
descriptions (e.g. governmental surveys, indigenous land studies)
where institutional independence and data sovereignty matter.
Pull the model first:
ollama pull gemma4:e4b # ~3 GB edge variant (default)
ollama pull gemma4:31b # frontier dense variant
ollama pull gemma3:27b # earlier generation, still solid
Then start an Ollama server (ollama serve) and the chat
object returned here will dispatch over HTTP locally.
Examples
if (FALSE) { # \dontrun{
# Cloud provider (needs ANTHROPIC_API_KEY)
provider <- vlm_provider("anthropic")
# Local Gemma 4 edge model -- default, ~3 GB, runs anywhere
provider <- vlm_provider("ollama")
# Local Gemma 4 frontier dense model -- best quality
provider <- vlm_provider("ollama", model = "gemma4:31b")
# Any other multimodal model the user has pulled
provider <- vlm_provider("ollama", model = "qwen2.5vl:32b")
} # }