async completion in pure vim script for vim8 and neovim
Vim script
Switch branches/tags
Nothing to show
Clone or download


Provide async autocompletion for vim8 and neovim with timers. This repository is fork of in pure vim script with python dependency removed.


Plug 'prabirshrestha/asyncomplete.vim'

Tab completion

inoremap <expr> <Tab> pumvisible() ? "\<C-n>" : "\<Tab>"
inoremap <expr> <S-Tab> pumvisible() ? "\<C-p>" : "\<S-Tab>"
inoremap <expr> <cr> pumvisible() ? "\<C-y>" : "\<cr>"

Force refresh completion

imap <c-space> <Plug>(asyncomplete_force_refresh)

Auto popup

By default asyncomplete will automatically show the autocomplete popup menu as you start typing. If you would like to disable the default behvior set g:asyncomplete_auto_popup to 0.

let g:asyncomplete_auto_popup = 0

You can use the above <Plug>(asyncomplete_force_refresh) to show the popup or can you tab to show the autocomplete.

let g:asyncomplete_auto_popup = 0

function! s:check_back_space() abort
    let col = col('.') - 1
    return !col || getline('.')[col - 1]  =~ '\s'

inoremap <silent><expr> <TAB>
  \ pumvisible() ? "\<C-n>" :
  \ <SID>check_back_space() ? "\<TAB>" :
  \ asyncomplete#force_refresh()
inoremap <expr><S-TAB> pumvisible() ? "\<C-p>" : "\<C-h>"

Remove duplicates

If you have many sources enabled (especially the buffer source), it might be useful to remove duplicates from the completion list. You can enable this by setting g:asyncomplete_remove_duplicates to 1.

let g:asyncomplete_remove_duplicates = 1

Preview Window

To disable preview window:

set completeopt-=preview

To enable preview window:

set completeopt+=preview

To auto close preview window when completion is done.

autocmd! CompleteDone * if pumvisible() == 0 | pclose | endif


asyncomplete.vim deliberately does not contain any sources. Please use one of the following sources or create your own.

Language Server Protocol (LSP)

Language Server Protocol via vim-lsp and asyncomplete-lsp.vim

Please note that vim-lsp setup for neovim requires neovim v0.2.0 or higher, since it uses lambda setup.

Plug 'prabirshrestha/asyncomplete.vim'
Plug 'prabirshrestha/async.vim'
Plug 'prabirshrestha/vim-lsp'
Plug 'prabirshrestha/asyncomplete-lsp.vim'

if executable('pyls')
    " pip install python-language-server
    au User lsp_setup call lsp#register_server({
        \ 'name': 'pyls',
        \ 'cmd': {server_info->['pyls']},
        \ 'whitelist': ['python'],
        \ })

Refer to vim-lsp wiki for configuring other language servers. Besides auto-complete language server support other features such as go to definition, find references, renaming symbols, document symbols, find workspace symbols, formatting and so on.

in alphabetical order

Languages/FileType/Source Links
Buffer asyncomplete-buffer.vim
Emoji asyncomplete-emoji.vim
Filenames / directories asyncomplete-file.vim
Go asyncomplete-gocode.vim
JavaScript (Flow) asyncomplete-flow.vim
Neosnippet asyncomplete-neosnippet.vim
Omni asyncomplete-omni.vim
PivotalTracker stories asyncomplete-pivotaltracker.vim
Racer asyncomplete-racer.vim
tmux complete tmux-complete.vim
Typescript asyncomplete-tscompletejob.vim
UltiSnips asyncomplete-ultisnips.vim
Vim Syntax asyncomplete-necosyntax.vim
Vim tags asyncomplete-tags.vim
Vim asyncomplete-necovim.vim

can't find what you are looking for? write one instead an send a PR to be included here or search github topics tagged with asyncomplete at

Using existing vim plugin sources

Rather than writing your own completion source from scratch you could also suggests other plugin authors to provide a async completion api that works for asyncomplete.vim or any other async autocomplete libraries without taking a dependency on asyncomplete.vim. The plugin can provide a function that takes a callback which returns the list of candidates and the startcol from where it must show the popup. Candidates can be list of words or vim's complete-items.

function s:completor(opt, ctx)
  call mylanguage#get_async_completions({candidates, startcol -> asyncomplete#complete(a:opt['name'], a:ctx, startcol, candidates) })

au User asyncomplete_setup call asyncomplete#register_source({
    \ 'name': 'mylanguage',
    \ 'whitelist': [*],
    \ 'completor': function('s:completor'),
    \ })


Use priority to control the order of the source. Highest priority comes first. priority is optional and defaults to 0 when registering a source.


function! s:js_completor(opt, ctx) abort
    let l:col = a:ctx['col']
    let l:typed = a:ctx['typed']

    let l:kw = matchstr(l:typed, '\v\S+$')
    let l:kwlen = len(l:kw)

    let l:startcol = l:col - l:kwlen

    let l:matches = [
        \ "do", "if", "in", "for", "let", "new", "try", "var", "case", "else", "enum", "eval", "null", "this", "true",
        \ "void", "with", "await", "break", "catch", "class", "const", "false", "super", "throw", "while", "yield",
        \ "delete", "export", "import", "public", "return", "static", "switch", "typeof", "default", "extends",
        \ "finally", "package", "private", "continue", "debugger", "function", "arguments", "interface", "protected",
        \ "implements", "instanceof"
        \ ]

    call asyncomplete#complete(a:opt['name'], a:ctx, l:startcol, l:matches)

au User asyncomplete_setup call asyncomplete#register_source({
    \ 'name': 'javascript',
    \ 'whitelist': ['javascript'],
    \ 'priority': 5,
    \ 'completor': function('s:js_completor'),
    \ })

The above sample shows synchronous completion. If you would like to make it async just call asyncomplete#complete whenever you have the results ready.

call timer_start(2000, {timer-> asyncomplete#complete(a:opt['name'], a:ctx, l:startcol, l:matches)})

If you are returning incomplete results and would like to trigger completion on the next keypress pass 1 as the fifth parameter to asyncomplete#complete which signifies the result is incomplete.

call asyncomplete#complete(a:opt['name'], a:ctx, l:startcol, l:matches, 1)

As a source author you do not have to worry about synchronization issues in case the server returns the async completion after the user has typed more characters. asyncomplete.vim uses partial caching as well as ignores if the context changes when calling asyncomplete#complete. This is one of the core reason why the original context must be passed when calling asyncomplete#complete.


All the credit goes to the following projects