The hardware and bandwidth for this mirror is donated by dogado GmbH, the Webhosting and Full Service-Cloud Provider. Check out our Wordpress Tutorial.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]dogado.de.
knitr::opts_chunk$set(
collapse = TRUE, comment = "#>",
eval = identical(tolower(Sys.getenv("LLMR_RUN_VIGNETTES", "false")), "true") )We will compare three configurations on two prompts, once unstructured and once with structured output. In choosing models, note that at the time of writing this vignette, Gemini models are not guaranteeing the schema output and is more likely to run into trouble.
library(LLMR)
library(dplyr)
cfg_openai <- llm_config("openai", "gpt-5-nano")
cfg_cld <- llm_config("anthropic","claude-sonnet-4-20250514", max_tokens = 512)
cfg_gem <- llm_config("groq", "openai/gpt-oss-20b")experiments <- build_factorial_experiments(
configs = list(cfg_openai, cfg_cld, cfg_gem),
user_prompts = c("Summarize in one sentence: The Apollo program.",
"List two benefits of green tea."),
system_prompts = c("Be concise.")
)
experiments
#> # A tibble: 6 × 6
#> config messages config_label user_prompt_label system_prompt_label
#> <list> <list> <chr> <chr> <chr>
#> 1 <llm_cnfg> <chr [2]> openai_gpt-5-nano user_1 system_1
#> 2 <llm_cnfg> <chr [2]> openai_gpt-5-nano user_2 system_1
#> 3 <llm_cnfg> <chr [2]> anthropic_claude-s… user_1 system_1
#> 4 <llm_cnfg> <chr [2]> anthropic_claude-s… user_2 system_1
#> 5 <llm_cnfg> <chr [2]> groq_openai/gpt-os… user_1 system_1
#> 6 <llm_cnfg> <chr [2]> groq_openai/gpt-os… user_2 system_1
#> # ℹ 1 more variable: repetition <int>setup_llm_parallel(workers = 10)
res_unstructured <- call_llm_par(experiments, progress = TRUE)
#> [2025-08-26 01:03:04.059187] LLMR Error: LLM API request failed.
#> HTTP status: 529
#> Reason: Overloaded
#> Tip: check model params for provider/API version.
#> [2025-08-26 01:03:04.243258] LLMR Error: LLM API request failed.
#> HTTP status: 529
#> Reason: Overloaded
#> Tip: check model params for provider/API version.
reset_llm_parallel()
res_unstructured |>
select(provider, model, user_prompt_label, response_text, finish_reason) |>
head()
#> # A tibble: 6 × 5
#> provider model user_prompt_label response_text finish_reason
#> <chr> <chr> <chr> <chr> <chr>
#> 1 openai gpt-5-nano user_1 "NASA's Apol… stop
#> 2 openai gpt-5-nano user_2 "- Rich in a… stop
#> 3 anthropic claude-sonnet-4-20250… user_1 <NA> error:server
#> 4 anthropic claude-sonnet-4-20250… user_2 <NA> error:server
#> 5 groq openai/gpt-oss-20b user_1 "The Apollo … stop
#> 6 groq openai/gpt-oss-20b user_2 "- **Rich in… stopschema <- list(
type = "object",
properties = list(
answer = list(type="string"),
keywords = list(type="array", items = list(type="string"))
),
required = list("answer","keywords"),
additionalProperties = FALSE
)
experiments2 <- experiments
experiments2$config <- lapply(experiments2$config, enable_structured_output, schema = schema)
setup_llm_parallel(workers = 10)
res_structured <- call_llm_par_structured(experiments2 , .fields = c("answer","keywords") )
reset_llm_parallel()
res_structured |>
select(provider, model, user_prompt_label, structured_ok, answer) |>
head()
#> # A tibble: 6 × 5
#> provider model user_prompt_label structured_ok answer
#> <chr> <chr> <chr> <lgl> <chr>
#> 1 openai gpt-5-nano user_1 TRUE "The Apoll…
#> 2 openai gpt-5-nano user_2 TRUE "1) Rich i…
#> 3 anthropic claude-sonnet-4-20250514 user_1 TRUE "The Apoll…
#> 4 anthropic claude-sonnet-4-20250514 user_2 TRUE "Green tea…
#> 5 groq openai/gpt-oss-20b user_1 TRUE "The Apoll…
#> 6 groq openai/gpt-oss-20b user_2 TRUE "Green tea…These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
Health stats visible at Monitor.