/get-query

Retrieves content recommendations for a given query. Works on any query, whether or not it was created via the API.

• Method: POST
• Endpoint: https://app.neuronwriter.com/neuron-api/0.5/writer/get-query

After creating a query with /new-query it usually takes ~60 seconds to prepare recommendations. Use /get-query to check whether analysis finished (status == "ready") and, when it has, read the recommendations from the response.

Parameters

Parameter	Example	Description
query	32dee2a89374a722	The ID of your query.

Response

Key	Description
status	Whether the query has been processed. Values: not found, waiting, in progress, ready. When ready, the keys below are included.
metrics	General recommendations such as typical content length and readability target. Example: {'word_count': {'median': 1864, 'target': 1864}, 'readability': {'median': 40, 'target': 40}}
terms_txt	Content term suggestions (title, desc_title, h1, h2, content_basic, content_extended) in plain text. Handy for feeding directly into ChatGPT/Claude prompts.
terms	Detailed term data: title/desc/h1/h2/content_basic/content_extended each as a list of {t, usage_pc, sugg_usage}, plus entities with {t, importance, relevance, confidence, links}.
ideas	Questions related to the topic: suggest_questions, people_also_ask, content_questions.
competitors	Basic info about the SERP competitors: rank, url, title, desc.
serp_summary	User intent + content-type stats across the TOP30 results. Example: {'top_intent': 'informational', 'intent_stats': {...}, 'top_content_type': 'educational', 'content_type_stats': {...}} (available for queries created after 2025-08-26).

Recommended polling pattern

import json, requests, time

API_ENDPOINT = 'https://app.neuronwriter.com/neuron-api/0.5/writer'
API_KEY = '<your-neuron-api-key>'
headers = {"X-API-KEY": API_KEY, "Accept": "application/json", "Content-Type": "application/json"}

def get_query(query_id):
    payload = json.dumps({"query": query_id})
    r = requests.post(API_ENDPOINT + "/get-query", headers=headers, data=payload)
    return r.json()

query_id = "79ca6b6b45fb9d67"
while True:
    data = get_query(query_id)
    if data.get("status") == "ready":
        break
    time.sleep(15)

print("Word count target:", data["metrics"]["word_count"]["target"])
print("Title terms:", data["terms_txt"]["title"])
print("PAA:", [q["q"] for q in data["ideas"]["people_also_ask"]])

For prompt-stuffing

terms_txt is the fastest path into an LLM prompt. The structured terms object is better when you want to programmatically check usage ranges and build a coverage score.