# HELP openai_requests_total OpenAI Realtime API requests counted via response.created
# TYPE openai_requests_total counter
openai_requests_total 0.0
# HELP user_requests_total Total number of user requests (audio + text)
# TYPE user_requests_total counter
user_requests_total 10.0
# HELP agent_left_listening_total Total number of times the agent left the listening state
# TYPE agent_left_listening_total counter
agent_left_listening_total 9.0
# HELP app_errors_total Total number of application errors
# TYPE app_errors_total counter
app_errors_total 1.0
# HELP llm_input_tokens_total Total input tokens sent to LLM (text + audio)
# TYPE llm_input_tokens_total counter
llm_input_tokens_total{model="gpt-realtime-2025-08-28"} 34984.0
# HELP llm_input_text_tokens_total Total input text tokens sent to LLM
# TYPE llm_input_text_tokens_total counter
llm_input_text_tokens_total{model="gpt-realtime-2025-08-28"} 34437.0
# HELP llm_input_audio_tokens_total Total input audio tokens sent to LLM
# TYPE llm_input_audio_tokens_total counter
llm_input_audio_tokens_total{model="gpt-realtime-2025-08-28"} 547.0
# HELP llm_input_cached_tokens_total Total cached input tokens reused by LLM
# TYPE llm_input_cached_tokens_total counter
llm_input_cached_tokens_total{model="gpt-realtime-2025-08-28"} 24960.0
# HELP llm_input_image_tokens_total Total input image tokens sent to LLM
# TYPE llm_input_image_tokens_total counter
llm_input_image_tokens_total{model="gpt-realtime-2025-08-28"} 0.0
# HELP llm_output_tokens_total Total output tokens generated by LLM (text + audio)
# TYPE llm_output_tokens_total counter
llm_output_tokens_total{model="gpt-realtime-2025-08-28"} 271.0
# HELP llm_output_text_tokens_total Total output text tokens generated by LLM
# TYPE llm_output_text_tokens_total counter
llm_output_text_tokens_total{model="gpt-realtime-2025-08-28"} 271.0
# HELP llm_output_audio_tokens_total Total output audio tokens generated by LLM
# TYPE llm_output_audio_tokens_total counter
llm_output_audio_tokens_total{model="gpt-realtime-2025-08-28"} 0.0
# HELP lk_agents_child_process_count Total number of child processes
# TYPE lk_agents_child_process_count gauge
lk_agents_child_process_count{nodename="enluma-ml-dev",pid="35999"} 0.0
# HELP total_conversational_latency_seconds Delay between user stopping speaking and agent beginning to speak
# TYPE total_conversational_latency_seconds histogram
total_conversational_latency_seconds_sum 2.1720123291015625
total_conversational_latency_seconds_bucket{le="0.05"} 0.0
total_conversational_latency_seconds_bucket{le="0.1"} 0.0
total_conversational_latency_seconds_bucket{le="0.2"} 0.0
total_conversational_latency_seconds_bucket{le="0.5"} 0.0
total_conversational_latency_seconds_bucket{le="1.0"} 3.0
total_conversational_latency_seconds_bucket{le="2.0"} 3.0
total_conversational_latency_seconds_bucket{le="5.0"} 3.0
total_conversational_latency_seconds_bucket{le="10.0"} 3.0
total_conversational_latency_seconds_bucket{le="+Inf"} 3.0
total_conversational_latency_seconds_count 3.0
# HELP agent_eou_delay_seconds EOU delay after accounting for LLM TTFT and TTS TTFB
# TYPE agent_eou_delay_seconds histogram
agent_eou_delay_seconds_sum 0.7419373570010066
agent_eou_delay_seconds_bucket{le="0.05"} 1.0
agent_eou_delay_seconds_bucket{le="0.1"} 1.0
agent_eou_delay_seconds_bucket{le="0.2"} 1.0
agent_eou_delay_seconds_bucket{le="0.5"} 3.0
agent_eou_delay_seconds_bucket{le="1.0"} 3.0
agent_eou_delay_seconds_bucket{le="2.0"} 3.0
agent_eou_delay_seconds_bucket{le="5.0"} 3.0
agent_eou_delay_seconds_bucket{le="10.0"} 3.0
agent_eou_delay_seconds_bucket{le="+Inf"} 3.0
agent_eou_delay_seconds_count 3.0
# HELP llm_ttft_seconds Time to first token for LLM responses
# TYPE llm_ttft_seconds histogram
llm_ttft_seconds_sum{model="gpt-realtime-2025-08-28"} 1.3253276348114014
llm_ttft_seconds_bucket{le="0.05",model="gpt-realtime-2025-08-28"} 0.0
llm_ttft_seconds_bucket{le="0.1",model="gpt-realtime-2025-08-28"} 0.0
llm_ttft_seconds_bucket{le="0.2",model="gpt-realtime-2025-08-28"} 2.0
llm_ttft_seconds_bucket{le="0.5",model="gpt-realtime-2025-08-28"} 6.0
llm_ttft_seconds_bucket{le="1.0",model="gpt-realtime-2025-08-28"} 6.0
llm_ttft_seconds_bucket{le="2.0",model="gpt-realtime-2025-08-28"} 6.0
llm_ttft_seconds_bucket{le="5.0",model="gpt-realtime-2025-08-28"} 6.0
llm_ttft_seconds_bucket{le="10.0",model="gpt-realtime-2025-08-28"} 6.0
llm_ttft_seconds_bucket{le="+Inf",model="gpt-realtime-2025-08-28"} 6.0
llm_ttft_seconds_count{model="gpt-realtime-2025-08-28"} 6.0
# HELP tts_ttfb_seconds Time to first byte for TTS responses
# TYPE tts_ttfb_seconds histogram
tts_ttfb_seconds_sum{model="gpt-realtime-2025-08-28"} 1.1600265000015497
tts_ttfb_seconds_bucket{le="0.05",model="gpt-realtime-2025-08-28"} 0.0
tts_ttfb_seconds_bucket{le="0.1",model="gpt-realtime-2025-08-28"} 0.0
tts_ttfb_seconds_bucket{le="0.2",model="gpt-realtime-2025-08-28"} 0.0
tts_ttfb_seconds_bucket{le="0.5",model="gpt-realtime-2025-08-28"} 5.0
tts_ttfb_seconds_bucket{le="1.0",model="gpt-realtime-2025-08-28"} 5.0
tts_ttfb_seconds_bucket{le="2.0",model="gpt-realtime-2025-08-28"} 5.0
tts_ttfb_seconds_bucket{le="5.0",model="gpt-realtime-2025-08-28"} 5.0
tts_ttfb_seconds_bucket{le="10.0",model="gpt-realtime-2025-08-28"} 5.0
tts_ttfb_seconds_bucket{le="+Inf",model="gpt-realtime-2025-08-28"} 5.0
tts_ttfb_seconds_count{model="gpt-realtime-2025-08-28"} 5.0