1~
2AI Agent da chuyen tu cac ban demo nghien cuu sang he thong production. Du kien hon 60% ung dung AI doanh nghiep se bao gom cac thanh phan agentic vao nam 2026. Nhung viec xay dung Agent tu dau  -  quan ly tool loop, state, memory, xu ly loi va phoi hop multi-agent  -  rat phuc tap. Do la ly do framework ra doi.
3~
4Bon framework thong tri nam 2026: **LangGraph**, **CrewAI**, **OpenAI Agents SDK** va **Claude Agent SDK**. Moi framework co cach tiep can hoan toan khac nhau cho cung mot van de: trao cho LLM kha nang suy luan, lap ke hoach, su dung cong cu va hop tac.
5~
6## Tong quan
7~
8| Khia canh | LangGraph | CrewAI | OpenAI Agents SDK | Claude Agent SDK |
9|--------|-----------|--------|-------------------|-----------------|
10| **Nha phat trien** | LangChain | CrewAI Inc. | OpenAI | Anthropic |
11| **Kien truc** | Dua tren do thi | Dua tren vai tro | Dua tren ban giao | Vong lap tu chu |
12| **Triet ly** | Kiem soat toi da | Hop tac nhom | Truu tuong toi thieu | Cho Agent mot may tinh |
13| **Ngon ngu** | Python, TypeScript | Python | Python | Python, TypeScript |
14| **Ho tro model** | Bat ky (OpenAI, Claude, local) | Bat ky | Bat ky (bat chap ten goi) | Chi Claude |
15| **GitHub stars** | ~29k | ~40k | ~21k | ~6k |
16| **Phu hop nhat cho** | Workflow phuc tap co state | Chuyen mon hoa multi-agent | Routing va triage | Lap trinh va tac vu nang ve file |
17~
18## LangGraph: Nha xay dung do thi
19~
20LangGraph mo hinh hoa workflow cua Agent duoi dang **do thi co huong co chu ky**. Ban dinh nghia cac node (ham thuc hien cong viec) va edge (chuyen tiep giua chung, co the la co dieu kien). State chay qua do thi va duoc luu tru thong qua checkpointing.
21~
22Day la framework ro rang va kiem soat duoc nhat  -  ban tu tay noi tung buoc.
23~
24```mermaid
25graph LR
26    Start --> Router[Router Node]
27    Router -->|needs research| Research[Research Node]
28    Router -->|needs code| Code[Code Node]
29    Research --> Synthesize[Synthesize Node]
30    Code --> Synthesize
31    Synthesize --> End
32```
33~
34### Khai niem cot loi
35~
36- **StateGraph**: dinh nghia do thi voi state co kieu
37- **Nodes**: ham Python chuyen doi state
38- **Edges**: ket noi giua cac node, co the co dieu kien
39- **Checkpointing**: luu tru tich hop cho workflow chay dai
40~
41### Vi du code
42~
43```python
44from langgraph.graph import StateGraph, MessagesState, START, END
45from langchain_openai import ChatOpenAI
46~
47llm = ChatOpenAI(model="gpt-4o")
48~
49def call_agent(state: MessagesState):
50    response = llm.invoke(state["messages"])
51    return {"messages": [response]}
52~
53def should_continue(state: MessagesState):
54    last = state["messages"][-1]
55    if last.tool_calls:
56        return "tools"
57    return END
58~
59def call_tools(state: MessagesState):
60    # Execute tool calls and return results
61    results = []
62    for tool_call in state["messages"][-1].tool_calls:
63        result = execute_tool(tool_call)
64        results.append(result)
65    return {"messages": results}
66~
67graph = StateGraph(MessagesState)
68graph.add_node("agent", call_agent)
69graph.add_node("tools", call_tools)
70graph.add_edge(START, "agent")
71graph.add_conditional_edges("agent", should_continue, {"tools": "tools", END: END})
72graph.add_edge("tools", "agent")
73~
74app = graph.compile()
75result = app.invoke({"messages": [{"role": "user", "content": "What's the weather?"}]})
76```
77~
78### Diem manh
79~
80- Kiem soat chi tiet tung buoc va tung chuyen tiep
81- Checkpointing va human-in-the-loop tich hop
82- Ho tro TypeScript day du
83- Hoat dong voi bat ky LLM provider nao
84- Phu hop nhat cho workflow phuc tap co phan nhanh dieu kien va vong lap
85~
86### Diem yeu
87~
88- Duong cong hoc tap doc  -  can hieu cac khai niem ly thuyet do thi
89- Dai dong cho use case don gian  -  agent co ban can nhieu boilerplate hon cac framework khac
90- Debug luong do thi co the kho khan neu khong co LangSmith
91~
92### Gia ca
93~
94Open-source (MIT). LangSmith (nen tang observability quan ly) co cac goi tra phi cho giam sat production.
95~
96## CrewAI: Nha to chuc nhom
97~
98CrewAI su dung phep an du cua con nguoi: ban tap hop mot **crew** gom cac agent chuyen biet, moi agent co **role**, **goal** va **backstory**. Cac agent hop tac tren cac **task** su dung **tool**, duoc dieu phoi boi **process** (tuan tu, phan cap hoac dong thuan).
99~
100Hay nghi ve no nhu viec tuyen mot doi ngu ma moi thanh vien co mot chuc danh va chuyen mon cu the.
101~
102```mermaid
103graph TD
104    Crew[Crew Manager] --> R[Researcher\nRole: Find data\nTools: WebSearch]
105    Crew --> W[Writer\nRole: Write content\nTools: FileWrite]
106    Crew --> E[Editor\nRole: Review quality\nTools: FileRead]
107    R --> Task1[Research Task]
108    W --> Task2[Writing Task]
109    E --> Task3[Review Task]
110    Task1 --> Task2 --> Task3
111```
112~
113### Khai niem cot loi
114~
115- **Agent**: persona co role, goal, backstory va tool
116- **Task**: nhiem vu co mo ta, dau ra ky vong va agent duoc chi dinh
117- **Crew**: nhom agent lam viec cung nhau
118- **Process**: chien luoc thuc thi (tuan tu, phan cap, dong thuan)
119- **Flow**: lop orchestration huong su kien de ket noi nhieu crew
120~
121### Vi du code
122~
123```python
124from crewai import Agent, Task, Crew, Process
125~
126researcher = Agent(
127    role="Senior Research Analyst",
128    goal="Find comprehensive data about the given topic",
129    backstory="You have 10 years of experience in technology research. "
130              "You are thorough and always verify facts from multiple sources.",
131    tools=[web_search_tool],
132    verbose=True,
133)
134~
135writer = Agent(
136    role="Technical Writer",
137    goal="Create clear, engaging technical content",
138    backstory="You write for a developer audience. "
139              "Your articles are practical and include code examples.",
140    tools=[file_tool],
141    verbose=True,
142)
143~
144research_task = Task(
145    description="Research the latest developments in WebAssembly in 2026. "
146                "Focus on WASI, Component Model, and production use cases.",
147    expected_output="A structured research document with key findings and sources.",
148    agent=researcher,
149)
150~
151writing_task = Task(
152    description="Write a blog post based on the research. "
153                "Include code examples and Mermaid diagrams.",
154    expected_output="A complete blog post in Markdown format.",
155    agent=writer,
156    context=[research_task],  # Writer receives researcher's output
157)
158~
159crew = Crew(
160    agents=[researcher, writer],
161    tasks=[research_task, writing_task],
162    process=Process.sequential,
163    verbose=True,
164)
165~
166result = crew.kickoff()
167print(result.raw)
168```
169~
170### Diem manh
171~
172- Truu tuong dua tren vai tro truc quan  -  de suy luan
173- Hon 100 tool integration tich hop
174- Bo nho chia se giua cac agent (ngan han, dai han, entity)
175- Cong dong lon nhat (~40k GitHub stars)
176- Process phan cap voi agent "quan ly" uy quyen va xac nhan
177~
178### Diem yeu
179~
180- Kiem soat it chi tiet hon LangGraph  -  ban dinh nghia vai tro, khong phai duong thuc thi chinh xac
181- Process phan cap co the kho du doan khi cac agent bat dong
182- Debug cuoc hoi thoai multi-agent kho hon luong single-agent
183~
184### Gia ca
185~
186Open-source core (mien phi). CrewAI Platform: $99/thang (Teams) den $120k/nam (Enterprise). Gia dua tren so crew hoat dong va so lan thuc thi hang thang.
187~
188## OpenAI Agents SDK: Bo dinh tuyen
189~
190OpenAI Agents SDK (nguoi ke thua tinh than cua Swarm) tap trung vao **handoff**  -  agent chuyen cuoc hoi thoai cho agent chuyen biet khac. Day la framework toi gian nhat: agent, tool, handoff va guardrail. Chi vay thoi.
191~
192```mermaid
193graph LR
194    User --> Triage[Triage Agent]
195    Triage -->|billing question| Billing[Billing Agent]
196    Triage -->|refund request| Refund[Refund Agent]
197    Triage -->|technical issue| Support[Support Agent]
198    Billing --> Response[Response]
199    Refund --> Response
200    Support --> Response
201```
202~
203### Khai niem cot loi
204~
205- **Agent**: model + instructions + tool + handoff
206- **Handoff**: chuyen tiep den agent khac (duoc mo hinh hoa nhu tool ma LLM co the goi)
207- **Guardrail**: xac thuc input/output chay song song voi agent
208- **Runner**: thuc thi agent loop
209- **Tracing**: observability tich hop cho moi cuoc goi LLM, tool va handoff
210~
211### Vi du code
212~
213```python
214from agents import Agent, Runner, handoff, InputGuardrail, GuardrailFunctionOutput
215from pydantic import BaseModel
216~
217class SafetyCheck(BaseModel):
218    is_safe: bool
219    reason: str
220~
221async def content_safety(ctx, agent, input_text):
222    result = await Runner.run(
223        Agent(name="Safety", instructions="Check if input is safe. No PII."),
224        input_text,
225        context=ctx,
226    )
227    output = SafetyCheck.model_validate_json(result.final_output)
228    return GuardrailFunctionOutput(
229        output_info=output, tripwire_triggered=not output.is_safe
230    )
231~
232billing_agent = Agent(
233    name="Billing Agent",
234    instructions="You handle billing inquiries. Be precise with numbers.",
235    tools=[lookup_invoice, process_payment],
236)
237~
238refund_agent = Agent(
239    name="Refund Agent",
240    instructions="You process refund requests. Always verify the order first.",
241    tools=[lookup_order, issue_refund],
242)
243~
244triage_agent = Agent(
245    name="Triage Agent",
246    instructions="Route the customer to the right specialist. "
247                 "Ask clarifying questions if needed.",
248    handoffs=[billing_agent, refund_agent],
249    input_guardrails=[InputGuardrail(guardrail_function=content_safety)],
250)
251~
252result = await Runner.run(triage_agent, "I need a refund for order #4521")
253print(result.final_output)
254# The triage agent routes to refund_agent, which processes the refund
255```
256~
257### Diem manh
258~
259- Mau handoff gon gang  -  tu nhien cho workflow routing/triage
260- Guardrail chay song song voi thuc thi (fail-fast, khong blocking)
261- Dashboard tracing tich hop de debug
262- Bat chap ten goi, ho tro cac model khong phai cua OpenAI
263- Truu tuong toi thieu  -  de hieu va de mo rong
264~
265### Diem yeu
266~
267- Quan ly state chua truong thanh bang LangGraph
268- Khong co persistence hoac checkpointing tich hop
269- Ecosystem tool cua ben thu ba nho hon
270- Thiet ke lay handoff lam trung tam co the khong phu hop moi kien truc
271~
272### Gia ca
273~
274Open-source (MIT). Ban tra phi theo token cho bat ky model nao su dung.
275~
276## Claude Agent SDK: Nha phat trien
277~
278Claude Agent SDK co cach tiep can khac: thay vi dinh nghia workflow hay role, ban cho agent **mot bo cong cu va de no tu tim cach hoan thanh nhiem vu**. No su dung cung vong lap tu chu nhu Claude Code  -  doc, hanh dong, xac minh, lap lai.
279~
280```mermaid
281graph TD
282    Prompt[User Prompt] --> Loop[Autonomous Agent Loop]
283    Loop --> Reason[Reason about next step]
284    Reason --> Act[Execute tool]
285    Act --> Verify[Check result]
286    Verify -->|not done| Loop
287    Verify -->|done| Output[Final output]
288```
289~
290### Khai niem cot loi
291~
292- **query()**: diem vao chinh khoi dong agent loop
293- **Cong cu tich hop**: Read, Write, Edit, Bash, Glob, Grep, WebSearch, WebFetch
294- **Custom tool qua MCP**: dinh nghia tool la in-process MCP server
295- **Sub-agent**: agent chuyen biet ma agent cha co the uy quyen
296- **Sessions**: duy tri context qua nhieu lan tuong tac
297~
298### Vi du code
299~
300```typescript
301import { tool, createSdkMcpServer, query } from "@anthropic-ai/claude-agent-sdk";
302import { z } from "zod";
303~
304const searchDocs = tool(
305  "search_docs",
306  "Search the internal documentation for relevant information",
307  { query: z.string().describe("Search query") },
308  async ({ query }) => {
309    const results = await vectorStore.similaritySearch(query, 5);
310    return {
311      content: [{ type: "text", text: results.map(r => r.pageContent).join("\n\n") }],
312    };
313  }
314);
315~
316const docsServer = createSdkMcpServer({
317  name: "docs",
318  version: "1.0.0",
319  tools: [searchDocs],
320});
321~
322for await (const message of query({
323  prompt: "Find how authentication works in our system and write a summary",
324  options: {
325    mcpServers: { docs: docsServer },
326    allowedTools: ["Read", "Glob", "Grep", "mcp__docs__search_docs"],
327  },
328})) {
329  if (message.type === "result" && message.subtype === "success") {
330    console.log(message.result);
331  }
332}
333```
334~
335### Diem manh
336~
337- Tich hop MCP cap first-class  -  ket noi voi bat ky ecosystem MCP server nao
338- Cong cu tich hop cho thao tac file, terminal va truy cap web
339- Tu dong nen context cho codebase lon
340- Sub-agent song song cho cac tac vu phuc tap
341- Cung engine voi Claude Code  -  da duoc thu nghiem thuc te tren workflow phat trien that
342~
343### Diem yeu
344~
345- Chi ho tro model Claude  -  khong ho tro nhieu provider
346- Framework moi hon voi cong dong nho hon
347- Can Node.js runtime ngay ca voi Python SDK
348- Kiem soat workflow it ro rang hon so voi LangGraph
349~
350### Gia ca
351~
352Open-source. Gia token Claude API tieu chuan. Managed Agents (phien ban hosted): $0.08 moi gio session ngoai chi phi token.
353~
354## Khi nao chon cai nao
355~
356```mermaid
357graph TD
358    Start{What's your priority?}
359    Start -->|Full control over workflow| LG[LangGraph]
360    Start -->|Multi-agent collaboration| CA[CrewAI]
361    Start -->|Routing and triage| OA[OpenAI Agents SDK]
362    Start -->|Coding and file automation| CS[Claude Agent SDK]
363~
364    LG --> LGU[Complex stateful workflows\nConditional branching\nHuman-in-the-loop]
365    CA --> CAU[Team of specialized agents\nResearch + writing pipelines\nContent generation]
366    OA --> OAU[Customer service routing\nMulti-step handoffs\nInput validation]
367    CS --> CSU[Code generation and review\nFile-heavy automation\nMCP tool ecosystem]
368```
369~
370### Chon LangGraph neu:
371- Ban can kiem soat chinh xac tung buoc cua workflow
372- Use case cua ban lien quan den logic dieu kien phuc tap va vong lap
373- Ban can persistence va human-in-the-loop checkpoint tich hop
374- Ban can su dung nhieu LLM provider trong cung mot workflow
375~
376### Chon CrewAI neu:
377- Ban muon truu tuong dua tren vai tro truc quan
378- Nhiem vu cua ban lien quan den nhieu agent co chuyen mon khac nhau
379- Ban can cac agent hop tac va truyen context cho nhau
380- Ban coi trong cong dong lon nhat va nhieu integration tich hop nhat
381~
382### Chon OpenAI Agents SDK neu:
383- Mau chinh cua ban la dinh tuyen cuoc hoi thoai den chuyen gia
384- Ban can guardrail xac thuc input/output song song
385- Ban muon truu tuong don gian nhat voi boilerplate toi thieu
386- Tracing va observability tich hop la quan trong
387~
388### Chon Claude Agent SDK neu:
389- Agent cua ban can doc, viet va thuc thi code
390- Ban muon tich hop MCP server cap first-class
391- Ban can agent tu chu co kha nang lap lai va tu sua loi
392- Ban da su dung Claude va muon tich hop sau nhat
393~
394## Co the ket hop cac framework khong?
395~
396Co. Mot mau pho bien la su dung mot framework cho orchestration va mot framework khac cho tung agent rieng le:
397~
398- **LangGraph** cho do thi workflow tong the
399- **CrewAI** cho node cu the can hop tac multi-agent
400- **Claude Agent SDK** cho sub-task lien quan den lap trinh qua MCP
401- **OpenAI Agents SDK** cho triage va routing huong toi khach hang
402~
403Cac framework nay khong loai tru lan nhau. Hay su dung cai phu hop voi tung phan cua he thong.
404~
405## Ket luan
406~
407Moi framework co mot dinh huong ro rang:
408~
409- **LangGraph** toi uu hoa cho kiem soat  -  ban quyet dinh tung chuyen tiep
410- **CrewAI** toi uu hoa cho hop tac  -  cac agent lam viec nhu mot doi
411- **OpenAI Agents SDK** toi uu hoa cho su don gian  -  truu tuong toi thieu, handoff gon gang
412- **Claude Agent SDK** toi uu hoa cho tu chu  -  cho no cong cu va de no lam viec
413~
414Lua chon dung phu thuoc vao workflow, doi ngu va stack cong nghe hien tai cua ban. Chon cai phu hop voi use case chinh, hoc ky no va dua cac framework khac vao khi ban gap the manh cua chung.
415~