Multi-Agent Orchestration: Full Benchmark Results
Notebook #18 — SynapseKit 1.4 vs LangChain 1.2 + LangGraph vs LlamaIndex Core 0.14
Total Lines of Code — 2-Agent Sequential Pipeline
SynapseKit
16
3 imports + 13 functional
LlamaIndex
19
3 imports + 16 functional
LangChain
23
4 imports + 19 functional
LoC Breakdown & Pattern Coverage
Orchestration Pattern Matrix
| Pattern |
SynapseKit |
LangChain |
LlamaIndex |
| Sequential |
Yes | Yes | Yes |
| Parallel |
Yes | Yes | No |
| Supervisor |
Yes | Yes | No |
| Handoff chain |
Yes | No | Yes |
| Graph / DAG |
Yes | Yes | No |
| Shared state |
Yes | Yes | Yes |
| Score (out of 6) |
6 |
5 |
3 |
Design Philosophy
SynapseKit
Task-centric
Define what each agent should produce. The framework manages context passing and execution order. Best for linear pipelines where the task dependency graph is simple.
LangChain (LangGraph)
Graph-centric
Define how data flows between nodes. Every routing decision is explicit code. Maximum flexibility for conditional workflows, retry loops, and parallel branches. LLM is just a node function.
LlamaIndex
Agent-centric
Agents decide when to hand off. Orchestration is emergent from LLM tool calls. Elegant for simple handoff chains. Non-deterministic — the LLM controls the transitions.
Benchmark Summary
| Criterion |
SynapseKit |
LangChain |
LlamaIndex |
| Least LoC (sequential) | 🏆 16 | 23 | 19 |
| Pattern coverage | 🏆 6/6 | 5/6 | 3/6 |
| Graph flexibility | Good | 🏆 Best | None |
| Deterministic routing | Yes | Yes | No (LLM-driven) |
| Intermediate step visibility | Partial | 🏆 Full | Partial |