Settings to use online mode for free

[benhaotang]

Following is a way to use the fast online mode completely free with a free gemini api key and a free jina api key

[LocalAI]
ollama_base_url = http://localhost:11434
default_model_ctx = -1
reason_model_ctx = -1

[API]
# your gemini key, get free at https://ai.google.dev/aistudio
openai_compat_api_key = AIxxxx
# your jina key, get free even without sign up at https://jina.ai/
jina_api_key = jina_xxxx
openai_url = https://generativelanguage.googleapis.com/v1beta/openai/chat/completions
jina_base_url = https://r.jina.ai/
searxng_url = http://localhost:4000/search

[Settings]
use_jina = true
use_ollama = false
with_planning = true
default_model = gemini-2.0-flash
reason_model = gemini-2.0-flash-thinking-exp-01-21

[Concurrency] # for browser
concurrent_limit = 3
cool_down = 10.0
chrome_port = 9222
chrome_host_ip = http://localhost
use_embed_browser = false

[Parsing]
temp_pdf_dir = ./temp_pdf
browse_lite = 0
pdf_max_pages = 30
pdf_max_filesize = 20971520
timeout_pdf = 75
max_html_length = 5120
max_eval_time = 15
verbose_web_parse_detail = false

[Ratelimits] # for AI endpoints
request_per_minute = 15
operation_wait_time = 0
fallback_model = gemini-2.0-flash-lite-preview-02-05