Tangyuan Toolbox

WeChat Mini Program Name Generator

Generate WeChat Mini Program name ideas from a category and up to three core keywords (space- or comma-separated, all included), with local validation, six-dimension scoring and search/brand/scenario commentary.

FreeOnline Tool
Loading…

Support this site

Always free. Tipping is entirely optional.

Customer Service

Click to view QQ number
877716587

How to Use

Choose a category (food, education, beauty and wellness, parenting, local services, or general), or leave it empty and enter only core keywords to use the general templates. You can enter up to three core keywords, each up to 6 characters; separate them with spaces. Commas, enumeration commas and slashes are also recognized and normalized to spaces, for example “bookkeeping reconciliation”. Every keyword appears in every generated name. Brand word (up to 8 characters), city/region and scene word are optional and feed the corresponding templates.

Four style switches control generation: reduplication and rhyme, persona prefixes such as “Old Wang’s”, city names in the result, and Chinese-only mode. Chinese-only mode rejects English letters and digits in the inputs and removes them from the results.

Click “Generate 10 names” to combine up to 2,000 candidates from category templates and weighted word slots. Each candidate goes through hard checks (width 4–30, protected words, long English/number runs, symbol or emoji at the start, four or more repeated characters, already-favorited names, approximate trademark prefixes) and six scoring dimensions: keyword position 25, readability 20, memorability 20, filler words 15, scalability 10 and phonetic friendliness 10. Multiple core keywords are inserted consecutively in input order, and each one’s position is scored. After exact and phonetic deduplication plus diversity quotas, 10 names are shown. If too few candidates survive, the tool relaxes the switches and tells you.

Results can be sorted by overall score, memorability or length, and filtered to risk-free names, names containing the city, or names containing the scene word. Each name shows three angles: search (are the core keywords near the front), brand (readable, memorable, no filler, scalable, smooth), and scenario (is the core keyword supported by a suitable scene or carrier such as “assistant”, “toolkit”, “station” or “ledger”). The scenario score is a surface-level word match and also affects ranking; contrast words raise memorability but lower the scenario score. When you enter a scene word, at least one candidate containing it is kept. “Details” shows the three-angle commentary, each dimension’s score and reason, the deduction list, hard-check issues and risk notes. When a fix is available (advertising filler, a city locked into a non-local category, repeated characters, an overlong English/number run), click “Apply suggestion” to replace the name and re-score it.

Favorites are stored in your browser’s localStorage and are skipped in later generations. The favorites bar copies a name on click and can be cleared. Batch scoring accepts up to 50 names, one per line, uses the same validation, six-dimension scoring and scenario score, and can copy the table as tab-separated text for Excel. Everything runs locally; no inputs are uploaded and no network requests are made.

Features

  • Six category template sets with weighted word slots; up to three core keywords all enter the candidates
  • Up to 2,000 combinations, validated before scoring so unusable names never reach the shortlist
  • Six explainable dimensions: keyword position, readability, memorability, filler words, scalability and phonetic friendliness
  • Search / brand / scenario commentary answering “findable?”, “memorable?” and “does the scenario hold?”
  • The scenario score also affects ranking so names that fit real usage surface first
  • A plain-language reason for every dimension and a line-by-line deduction list
  • Exact-character and phonetic deduplication to remove near-duplicates
  • Diversity quotas for templates, prefixes and repeated ending words
  • Four risk levels with text labels, not color alone
  • One-click rewrite suggestions for filler words, locked cities, repeated characters and long alphanumeric runs
  • Copy one name, copy all names, or copy names only
  • Local favorites that future generations avoid
  • Batch scoring for up to 50 names with TSV export
  • Fully local calculation: no uploads and no online duplicate checks

Use Cases

Naming before launch
Generate candidates by category and core keyword, filter out overly long, risky or awkward names, then take 10 names to the platform for availability checks.
Naming a local shop mini program
Pick food, beauty or local services, add a city and scene word, and generate friendly names that fit a neighborhood business.
Quick name health check
Run a batch of existing name ideas through the same validation and scoring to see hard-check issues, weak dimensions and rewrite suggestions.
Comparing brand candidates
Keep the core keyword and try different categories or style switches, then compare memorability, scalability and risk notes before deciding.

FAQ

Can I register a generated name directly?
Not guaranteed. The tool only combines, validates and scores names locally; it does not check duplicates or trademarks online. Before deciding, paste the name into the WeChat public platform to check availability and search official trademark records. The platform’s review result is final.
Do the scores represent WeChat search ranking?
No. Match score, memorability and phonetic friendliness are local heuristic scores based on keyword position, length, reduplication and tone patterns. They help compare candidates; they are not search rankings and do not guarantee approval.
How accurate is the trademark warning?
It is indicative only. The tool contains a small sample of common trademark prefixes and performs prefix matching to remind you to check officially. It is not a trademark search, does not cover all trademarks or classes, and should not be treated as a legal conclusion.
Why do some names include words like “Lab” or “Editorial Office”?
That is the contrast structure: pairing a concrete category word with an institutional word to create a memorable twist. If you do not want this style, turn off the style switches, generate another batch, or choose more conservative candidates.
How many core keywords can I enter? Will every one appear in the name?
Up to three keywords, each up to 6 characters, separated by spaces. Commas, enumeration commas and slashes are also recognized and normalized to spaces. The tool inserts all keywords consecutively in input order, so “bookkeeping reconciliation” produces names such as “bookkeeping reconciliation station” that contain both words. Keywords are the main recognition and search anchor, and none are dropped.
What do the search, brand and scenario angles measure?
Search measures how close the core keywords are to the front. Brand covers readability, memorability, advertising filler, scalability and phonetic smoothness. Scenario checks whether the core keyword is supported by a carrier such as assistant, toolkit, station or ledger, or by the scene word you entered. The scenario score is a local surface-level word match, not semantic understanding, and does not predict real-world performance.
Why does Chinese-only mode reject my input?
Chinese-only mode does not allow English letters or digits in the core keyword, brand word, city or scene word, and it removes them from generated names. Change the input to Chinese or turn the switch off.
Where are favorites stored?
In the current browser’s localStorage. They are never uploaded, and future generations skip favorited names. Clearing browser data or switching devices removes them.
How many names can I score at once?
Up to 50 names, one per line; duplicate lines are removed. Results include the original name, match score, memorability, main issue and a suggested rewrite, and can be copied as tab-separated text for Excel.
Are my keywords uploaded?
No. Name generation, pinyin calculation, validation, scoring and copying all run in your browser. The tool makes no network requests and does not collect your inputs.