Agent Orchestration

Multi-LLM coordination · Ollama local · Opik traced

OLLAMA MODEL:

→ http://localhost:11434

AGENT ROSTER

OpenAI GPT-4o

LATENCY

1240ms

TOKENS

84.2k

SUCCESS

98.4%

Gemini 1.5 Pro

LATENCY

980ms

TOKENS

62.1k

SUCCESS

96.1%

Claude 3.5 Sonnet

LATENCY

1500ms

TOKENS

53.8k

SUCCESS

97.2%

Ollama / Local LLM

LATENCY

420ms

TOKENS

31.1k

SUCCESS

94.3%

IBM Quantum

LATENCY

8200ms

TOKENS

—

SUCCESS

89%

Kimi

LATENCY

2100ms

TOKENS

18.6k

SUCCESS

93.7%

Manus

LATENCY

—

TOKENS

—

SUCCESS

—

PROMPT PLAYGROUND — MULTI-MODAL

OPIK TRACE VIEWER

Workspace: bosskali561-afid · Project: lspa-v3

[ Opik Dashboard ]

Every agent call is traced with @track decorator

Input · Output · Latency · Model · Phase · Cost