300-560 Network Programmability Developer Exam Guide
The 300-560 Network Programmability Developer (NPDEV) exam was designed to validate practical knowledge of Python-based network automation, data formats, controller platforms, device APIs, Cisco SDKs, and networking fundamentals. It is most relevant to candidates working from the historical NPDEV blueprint or checking an existing exam plan. The key decision now is whether 300-560 is still the correct target: Cisco’s current exam list does not display it, and Cisco says DevNet certifications moved to the CCNA, CCNP, and CCIE Automation track starting February 3, 2026.
Confirm that 300-560 is still the right exam
Check Cisco’s current exam catalogue and certification update before buying preparation material or scheduling anything. Cisco’s current exams list does not display 300-560, while its certification update says DevNet certifications moved to the CCNA, CCNP, and CCIE Automation track starting February 3, 2026. This means a candidate should verify an existing authorization, transition arrangement, or replacement exam directly with Cisco before continuing with a 300-560 study plan.
The historical exam overview identifies 300-560 as the Network Programmability Developer (NPDEV) exam. That identification is useful when matching older training, a previous study record, or an existing registration, but it should not be treated as proof that the exam is currently schedulable.
Use the official Cisco pages as the decision point: first confirm whether 300-560 can still be registered, then identify the current automation-track exam that matches your objective if it cannot. A practice site can organize preparation around the published blueprint, but it cannot replace Cisco’s live status and registration information.
What the certification transition changes
Cisco’s update describes a move from DevNet certifications to the CCNA, CCNP, and CCIE Automation track. The practical consequence is that a learner who wants a current Cisco automation credential should compare the new track with the historical NPDEV objectives rather than assume that older 300-560 material remains the correct endpoint.
If you already have a confirmed 300-560 appointment, check the appointment and Cisco’s current candidate guidance before changing plans. If you have only selected 300-560 as a future goal, pause scheduling and map your needs to the replacement automation path first.
What 300-560 was intended to validate
The 300-560 blueprint covers the developer skills needed to work across code, structured data, controllers, device interfaces, Cisco SDKs, and networking concepts. It is not limited to writing Python syntax: the blueprint connects programming decisions to network state, APIs, models, configuration deployment, and platform behavior.
A strong candidate profile combines software reasoning with network operations knowledge. You should be able to read a short Python program, identify how it handles a response, choose an appropriate API interaction, and recognize whether the resulting network operation is logically sound.
This makes the exam more suitable for network engineers moving into automation, developers learning Cisco networking, and automation practitioners who need Cisco-specific platform and protocol knowledge. The blueprint itself, rather than a generic Python syllabus, should determine how you divide preparation time.
The main capability areas
The official topics identify programming fundamentals, data handling and formats, network controller platforms and protocols, device programmability, APIs, Cisco SDKs, and networking fundamentals. Treat these as connected capabilities rather than isolated vocabulary lists.
For example, a RESTCONF task may require Python requests, JSON or XML handling, a YANG-based path, authentication awareness, and enough networking context to understand what the configuration changes. Studying each item separately without practicing the chain between them can leave gaps even when individual definitions look familiar.
How the historical exam was delivered
Cisco’s exam overview describes 300-560 as a 90-minute assessment containing 65–75 questions. The same overview lists English as the available language and identifies Pearson VUE as the registration and delivery provider. These details belong to the historical exam overview and should be rechecked if Cisco provides a current transition or replacement route.
The time limit creates a practical reading decision: do not spend the whole attempt proving one uncertain answer. Read the task, identify the technology and requested outcome, eliminate options that conflict with the protocol or data format, and move on when further analysis is not productive.
Do not infer a passing score, question format beyond the published description, retake rule, price, or current appointment availability from the blueprint. Those details are either time-sensitive or not supplied in the official research for this guide.
Which blueprint version applies
Cisco’s v1.1 exam-topics PDF says that version applies to written exams scheduled on June 26, 2018, and later. Cisco’s older v1.0 page recommends that version for candidates scheduled to take the written exam before June 26, 2018.
If you are reviewing archived material, establish the version before studying. Mixing v1.0 and v1.1 documents can create needless uncertainty about terminology and coverage. For any current registration question, use Cisco’s current candidate information rather than relying only on an old version note.
Build a study plan from the blueprint
Start with a skills inventory, not a list of downloadable questions. Mark each blueprint area as strong, familiar, or untested, then study the weakest prerequisite that blocks practical work. A useful sequence is Python and data handling first, followed by network protocols and controllers, then device APIs and integrated troubleshooting.
This order reflects dependency. You cannot reliably debug a RESTCONF script if you cannot distinguish a Python data structure from a JSON payload, and you cannot judge an API result if you do not understand the network object or configuration being changed.
Keep a short error log throughout preparation. Record the technology, the mistaken assumption, the evidence that disproved it, and the rule you will apply next time. This converts wrong answers into targeted review instead of repeated passive reading.
A four-stage roadmap
Stage one is programming and representation. Practice constructing, interpreting, and debugging Python code. Include examples that use JSON, XML, requests, and ncclient, because these are explicitly identified in the programming-fundamentals scope. Focus on data flow: input, transformation, request, response, and exception or validation behavior.
Stage two is network automation semantics. Review RESTCONF and NETCONF, YANG models, URI construction, configuration deployment, and the difference between a payload’s syntax and the network operation it expresses. Write small exercises that translate a model or intended change into the representation an API expects.
Stage three is platform coverage. Study controller concepts and device-specific APIs, then compare how the same automation objective differs between a controller and a directly managed device. Avoid learning only endpoint names; learn what resource, policy, or configuration object each interaction addresses.
Stage four is integration and review. Take a scenario, identify the platform, choose the protocol and data format, trace the request and response, and explain what should be checked before declaring success. Finish each session by updating your error log and revisiting one previously weak topic.
How to allocate study time without official weights
The supplied official research does not provide blueprint percentages, so do not assign unsupported weights to the domains. Instead, allocate time according to your diagnostic results and the number of dependencies in each topic. A candidate who can write Python but has never used YANG needs a different plan from a network engineer who understands APIs but cannot debug code.
Use two measures to decide when to move on: explanation and execution. You should be able to explain why an answer is appropriate and complete a small, controlled task that demonstrates the same principle. If you can memorize a definition but cannot trace a payload or identify a faulty request, the topic is not yet secure.
Master programming fundamentals before platform details
Programming fundamentals are the fastest way to expose whether your preparation is practical. The official scope includes constructing, interpreting, and debugging Python code, including code that uses JSON, XML, requests, and ncclient. Study these as tools for network work, not as a general software-development course.
Begin with variables, collections, conditions, loops, functions, exceptions, and imports. Then read short automation snippets line by line. For every snippet, state the input type, the transformation being performed, the request or library call, the expected response type, and the failure that would occur if a key or value were missing.
Next, practice modifying working code. Change a filter, alter a request body, handle a failed response, or extract a nested value. Small controlled changes teach more than copying a long script because they make the relationship between code and network data visible.
Common coding mistakes to eliminate
A frequent mistake is confusing a Python dictionary with a JSON document. A dictionary is an in-memory Python structure; JSON is a serialized data representation. Your preparation should include converting between representations and checking whether a library expects text, bytes, or a native object.
Another mistake is debugging only the syntax. A script can run and still send the wrong resource path, method, payload shape, or value type. When reviewing code, separate syntax errors, runtime errors, transport failures, authentication failures, response-format errors, and incorrect network intent.
With ncclient, requests, or similar libraries, pay attention to the boundary between the library call and the protocol operation. Know what the code is trying to send, what the response means, and which part of the result should be validated before proceeding.
Treat JSON, XML, and YANG as connected skills
The data-handling domain requires more than recognizing braces and tags. It includes constructing and parsing JSON or XML payloads, manipulating network entries, validating YANG models, and translating YANG models into RESTCONF URI and JSON representations. Prepare by tracing a data object from model to request to response.
For JSON, practice nested objects, arrays, key lookup, value updates, and serialization. For XML, practice namespaces, hierarchy, attributes, and element selection. The goal is not to memorize a single payload but to understand how structure affects parsing and how an incorrect structure can change the meaning of an operation.
For YANG, start with the model’s hierarchy and identity. Identify containers, lists, leaves, keys, and relationships, then use that structure to reason about a RESTCONF path and payload. Validate the intended object before writing the request so that you are not debugging a URI and a model error at the same time.
A repeatable payload exercise
Choose one simple configuration or operational-data task. First describe the desired network entry in plain language. Next locate the relevant model structure. Then sketch the URI or XML path, construct the payload, and decide what response or subsequent read would confirm the result.
Repeat the task with both JSON and XML when the protocol allows it. Compare the hierarchy rather than memorizing punctuation. Finally, deliberately introduce one error—a wrong key, missing namespace, invalid nesting, or incorrect value—and explain how you would detect it. This creates a troubleshooting habit that transfers to unfamiliar scenarios.
Cover controller platforms as operating models
The controller-platform domain includes ACI policy models, ACI application profiles, APIC-EM services and its Network Information Database, service tickets, OpenFlow, OpFlex, rollback procedures, and resource utilization. These topics require conceptual organization: know what the controller represents, what the client requests, and how the platform reports state.
For ACI, connect policy objects to the application profile and the resulting network intent. For APIC-EM, connect services, the Network Information Database, and service tickets to the workflow being automated. For OpenFlow and OpFlex, focus on the control relationship and the role of the protocol rather than treating the names as interchangeable API products.
Rollback and resource utilization are especially useful review anchors. Ask what should be restored, what state should be checked, and what resource signal could indicate that an automation action is unsafe or incomplete. These questions encourage operational reasoning instead of product-name recall.
How to compare a controller with a device API
A controller typically exposes an abstraction of policy, service, or topology, while a device API exposes configuration or operational resources on a particular platform. When studying a scenario, identify the source of truth, the object being changed, the scope of the operation, and how success is verified.
Do not assume that an identical intent produces an identical request. A controller may translate policy into device changes, whereas a device API may require direct configuration details. Your notes should show both the desired outcome and the layer at which the request is made.
Study device programmability by platform
The device-programmability domain covers NX-OS API, ASA API, and IOS XE/XR, including RESTCONF and NETCONF configuration deployment. Organize study by platform and task: identify the interface, understand the resource or configuration target, construct the request, and verify the result.
For NX-OS and IOS XE/XR, compare the API and model-driven approaches with the configuration or operational objective. For ASA, focus on the objects and security structures named in the blueprint rather than treating firewall automation as ordinary interface configuration.
The blueprint includes ASA network objects, access lists, security groups, and requests that combine multiple ASA REST operations in one POST. That combination deserves deliberate practice because it tests sequencing and payload organization, not merely recognition of an individual endpoint.
A device-API comparison checklist
For each platform, write down the authentication assumptions provided by a scenario, the request method or protocol, the data format, the resource path, the configuration effect, and the verification step. Then compare two platforms performing related tasks and note what cannot be carried over unchanged.
Review failure handling as part of the task. Consider an invalid model path, an incomplete payload, an authorization problem, or a successful HTTP response that does not prove the intended configuration is active. Good automation preparation includes checking state after the request.
Use virtual platforms for focused practice
Cisco lists VIRL, APIC Simulator, NX-OSv, DevNet sandboxes, ASAv, and CSR1000v as virtual platforms to compare for testing network applications. Use this list to select a practice environment that matches the skill you are learning, rather than trying to build every possible lab.
A lab is valuable when it answers a narrow question. For example, use an appropriate simulated environment to test a model-driven request, inspect a response, alter a payload, or compare controller and device behavior. Keep the code and payload in versioned notes so that each result can be reproduced or explained later.
If a platform is unavailable, do not pretend that reading a screenshot proves execution skill. Use documentation and static exercises for concepts, then mark the unverified behavior for later lab confirmation. This distinction prevents false confidence.
A practical lab sequence
Begin with read operations and harmless data extraction. Move to constructing a payload, submitting a controlled change, reading the resulting state, and handling an induced error. Only then combine multiple operations or platform concepts.
After each lab, write a short record: objective, platform, protocol, data format, request outline, observed response, and verification method. The record becomes a revision tool and exposes whether you understand the operation or merely followed instructions.
Use practice questions as diagnosis, not a substitute for study
Practice questions can reveal gaps in syntax, protocol behavior, or platform concepts, but memorizing answer patterns is a weak preparation method. Use questions only after studying the underlying objective, and explain why every incorrect option fails. Never rely on exam dumps, leaked questions, or claims that memorization guarantees a pass.
A good review cycle has three passes. First answer from your current knowledge. Second, consult the relevant official objective or technical reference and identify the governing principle. Third, restate the problem in your own words and solve it without looking at the answer.
Be cautious with any item that gives an unexplained endpoint, version-specific behavior, or ambiguous scenario. Record it as a research question rather than forcing a confident conclusion. Cisco states that exam-topic guidelines may change without notice and that related topics may also appear on a specific exam delivery.
Questions to ask after every miss
Was the problem caused by Python logic, data representation, protocol choice, platform behavior, networking knowledge, or careless reading? Which fact would let you eliminate the wrong options? Could the same principle appear with a different device or payload format?
If you cannot answer those questions, the review is incomplete. Return to a small code or lab exercise, then attempt a new question on the same concept. The aim is transfer, not recognition of a repeated sentence.
Avoid preparation traps that waste time
The most damaging trap is studying an outdated target without checking Cisco’s current status. The next is treating 300-560 as a Python-only exam. The blueprint spans controllers, APIs, Cisco SDKs, device programmability, data handling, and networking fundamentals, so narrow preparation leaves predictable blind spots.
A third trap is collecting product terms without building relationships between them. Write diagrams and workflows that connect a model, URI, payload, API call, response, and verification step. A fourth is doing labs without notes: unrecorded success is difficult to reproduce and difficult to explain under time pressure.
Finally, do not use unsupported blueprint percentages to justify your schedule. No domain percentages are supplied in the official research for this article. Let diagnostics, dependencies, and demonstrated performance—not invented weighting—determine your priorities.
Signs that you should delay scheduling
Delay the decision if you cannot explain the difference between a controller request and a device configuration request, cannot parse the data your code receives, or cannot identify how a YANG model informs a RESTCONF representation. Also pause if you have not verified that 300-560 is currently an available target.
Readiness is stronger when you can solve unfamiliar variations, debug a small script, construct or inspect a payload, and explain how the result would be validated. These are practical indicators, not an official passing standard.
Make the final review operational
In the final review period, stop expanding the topic list and consolidate the workflows that connect code to network behavior. Revisit your error log, rebuild one small exercise for each weak capability, and practice deciding what information matters before selecting an answer.
Use a compact revision sheet with sections for Python behavior, JSON and XML structure, YANG-to-RESTCONF translation, controller concepts, device APIs, ASA objects and operations, and networking fundamentals. For each entry, include a rule, a small example, and a failure condition.
Before any scheduling action, verify the current exam name, availability, version guidance, language, provider, and appointment instructions through Cisco. The historical overview lists English, Pearson VUE, 90-minute delivery, and 65–75 questions for 300-560, but current transition information may determine whether those details remain actionable.
Your next actions
First, open Cisco’s current exam list and certification update and decide whether you are pursuing 300-560 or a current automation-track exam. Second, if the historical blueprint remains relevant to your confirmed plan, complete a diagnostic across every named capability. Third, select one lab environment from Cisco’s listed virtual-platform options and begin with a small read-and-verify workflow.
After the first study cycle, review your error log rather than simply taking another mixed quiz. Replace weak memorization with a code trace, payload exercise, model translation, or controlled API task. Then recheck Cisco’s official registration and candidate information before committing to an appointment.
Conclusion
300-560 preparation is most useful when it is treated as network-automation problem solving: understand the Python, follow the data, identify the controller or device layer, construct the correct API interaction, and verify the resulting state. However, the immediate decision comes first. Because Cisco’s current materials no longer display 300-560 and describe a move to the Automation track, confirm the active certification route before investing further in a historical NPDEV plan. Use the official blueprint only for a verified legacy pathway or as background when mapping your skills to the current Cisco automation exams.
Related exams
- 500-210 exam — SP Optical Technology Field Engineer Representative
- 650-059 exam — Cisco Lifecycle Services Advanced Routing and Switching (LCSARS)
