Skip to content

sweepai/cursortab.nvim

BranchesTags
 
 

Repository files navigation

cursortab.nvim

A Neovim plugin that provides local edit completions and cursor predictions. Currently supports custom models and models form Zeta (Zed) and SweepAI.

Warning

This is an early-stage project. Expect bugs, incomplete features, and breaking changes. Make sure to regularly update the plugin.

Requirements

  • Go 1.24.2+ (for building the server component)
  • Neovim 0.8+ (for the plugin)

Installation

Using lazy.nvim

{
  "leonardcser/cursortab.nvim",
  -- version = "*",  -- Use latest tagged version for more stability
  build = "cd server && go build",
  config = function()
    require("cursortab").setup()
  end,
}

Using packer.nvim

use {
  "leonardcser/cursortab.nvim",
  -- tag = "*",  -- Use latest tagged version for more stability
  run = "cd server && go build",
  config = function()
    require("cursortab").setup()
  end
}

Configuration

require("cursortab").setup({
  enabled = true,
  log_level = "info",  -- "trace", "debug", "info", "warn", "error"
  state_dir = vim.fn.stdpath("state") .. "/cursortab",  -- Directory for runtime files (log, socket, pid)

  keymaps = {
    accept = "<Tab>",           -- Keymap to accept completion, or false to disable
    partial_accept = "<S-Tab>", -- Keymap to partially accept, or false to disable
    trigger = false,            -- Keymap to manually trigger completion, or false to disable
  },

  ui = {
    colors = {
      deletion = "#4f2f2f",      -- Background color for deletions
      addition = "#394f2f",      -- Background color for additions
      modification = "#282e38",  -- Background color for modifications
      completion = "#80899c",    -- Foreground color for completions
    },
    jump = {
      symbol = "",              -- Symbol shown for jump points
      text = " TAB ",            -- Text displayed after jump symbol
      show_distance = true,      -- Show line distance for off-screen jumps
      bg_color = "#373b45",      -- Jump text background color
      fg_color = "#bac1d1",      -- Jump text foreground color
    },
  },

  behavior = {
    idle_completion_delay = 50,  -- Delay in ms after idle to trigger completion (-1 to disable)
    text_change_debounce = 50,   -- Debounce in ms after text change to trigger completion (-1 to disable)
    max_visible_lines = 12,      -- Max visible lines per completion (0 to disable)
    cursor_prediction = {
      enabled = true,            -- Show jump indicators after completions
      auto_advance = true,       -- When no changes, show cursor jump to last line
      proximity_threshold = 2,   -- Min lines apart to show cursor jump (0 to disable)
    },
  },

  provider = {
    type = "inline",                      -- Provider: "inline", "fim", "sweep", "sweepapi", "zeta", or "copilot"
    url = "http://localhost:8000",        -- URL of the provider server
    api_key_env = "",                     -- Env var name for API key (e.g., "OPENAI_API_KEY")
    model = "",                           -- Model name
    temperature = 0.0,                    -- Sampling temperature
    max_tokens = 512,                     -- Max tokens to generate
    top_k = 50,                           -- Top-k sampling
    completion_timeout = 5000,            -- Timeout in ms for completion requests
    max_diff_history_tokens = 512,        -- Max tokens for diff history (0 = no limit)
    completion_path = "/v1/completions",  -- API endpoint path
    fim_tokens = {                        -- FIM tokens (for FIM provider)
      prefix = "<|fim_prefix|>",
      suffix = "<|fim_suffix|>",
      middle = "<|fim_middle|>",
    },
    privacy_mode = true,                  -- Don't send telemetry to provider
  },

  blink = {
    enabled = false,    -- Enable blink source
    ghost_text = true,  -- Show native ghost text alongside blink menu
  },

  debug = {
    immediate_shutdown = false,  -- Shutdown daemon immediately when no clients
  },
})

For detailed configuration documentation, see :help cursortab-config.

Providers

The plugin supports six AI provider backends: Inline, FIM, Sweep, Sweep API, Zeta, and Copilot.

Provider Hosted Multi-line Multi-edit Cursor Prediction Streaming Model
inline ✓ Any base model
fim ✓ ✓ Any FIM-capable
sweep ✓ ✓ ✓ ✓ sweep-next-edit-1.5b
sweepapi ✓ ✓ ✓ ✓ sweep-next-edit-7b
zeta ✓ ✓ ✓ ✓ zeta
copilot ✓ ✓ ✓ ✓ GitHub Copilot

Context Per Provider:

Context inline fim sweep zeta sweepapi copilot
Buffer content ✓ ✓ ✓ ✓ ✓
Edit history ✓ ✓ ✓
Previous file state ✓ ✓
LSP diagnostics ✓ ✓
Treesitter context ✓ ✓ ✓
Git diff context ✓ ✓ ✓
Recent files ✓
User actions ✓

Inline Provider (Default)

Details

End-of-line completion using OpenAI-compatible API endpoints. Works with any OpenAI-compatible /v1/completions endpoint.

Requirements:

  • An OpenAI-compatible completions endpoint

Example Configuration:

require("cursortab").setup({
  provider = {
    type = "inline",
    url = "http://localhost:8000",
  },
})

Example Setup:

# Using llama.cpp
llama-server -hf ggml-org/Qwen2.5-Coder-1.5B-Q8_0-GGUF --port 8000

FIM Provider

Details

Fill-in-the-Middle completion using standard FIM tokens. Uses both prefix (before cursor) and suffix (after cursor) context. Compatible with Qwen, DeepSeek-Coder, and similar models. Works with any OpenAI-compatible /v1/completions endpoint.

Requirements:

  • An OpenAI-compatible completions endpoint with a FIM-capable model

Example Configuration:

require("cursortab").setup({
  provider = {
    type = "fim",
    url = "http://localhost:8000",
  },
})

Example Setup:

# Using llama.cpp with Qwen2.5-Coder 1.5B
llama-server -hf ggml-org/Qwen2.5-Coder-1.5B-Q8_0-GGUF --port 8000

# Or with Qwen 2.5 Coder 14B + 0.5B draft for speculative decoding
llama-server \
    -hf ggml-org/Qwen2.5-Coder-14B-Q8_0-GGUF:q8_0 \
    -hfd ggml-org/Qwen2.5-Coder-0.5B-Q8_0-GGUF:q8_0 \
    --port 8012 \
    -b 1024 \
    -ub 1024 \
    --cache-reuse 256

Sweep Provider

Details

Sweep Next-Edit 1.5B model for fast, accurate next-edit predictions. Sends full file for small files, trimmed around cursor for large files.

Requirements:

  • vLLM or compatible inference server
  • Sweep Next-Edit model downloaded from Hugging Face

Example Configuration:

require("cursortab").setup({
  provider = {
    type = "sweep",
    url = "http://localhost:8000",
  },
})

Example Setup:

# Using llama.cpp
llama-server -hf sweepai/sweep-next-edit-1.5b-GGUF --port 8000

# Or with a local GGUF file
llama-server -m sweep-next-edit-1.5b.q8_0.v2.gguf --port 8000

Sweep API Provider

Details

Sweep's hosted API for Next-Edit predictions. No local model setup required.

Requirements:

  • Create an account at sweep.dev and get your API token
  • Set the SWEEPAPI_TOKEN environment variable with your token

Example Configuration:

# In your shell config (.bashrc, .zshrc, etc.)
export SWEEPAPI_TOKEN="your-api-token-here"
require("cursortab").setup({
  provider = {
    type = "sweepapi",
    api_key_env = "SWEEPAPI_TOKEN",
  },
})

Zeta Provider

Details

Zed's Zeta model - a Qwen2.5-Coder-7B fine-tuned for edit prediction using DPO and SFT.

Requirements:

  • vLLM or compatible inference server
  • Zeta model downloaded from Hugging Face

Example Configuration:

require("cursortab").setup({
  provider = {
    type = "zeta",
    url = "http://localhost:8000",
    model = "zeta",
  },
})

Example Setup:

# Using vLLM
vllm serve zed-industries/zeta --served-model-name zeta --port 8000

# See the HuggingFace page for optimized deployment options

Copilot Provider

Details

GitHub Copilot completions using the official copilot-language-server LSP server, enabled with vim.lsp.enable. Can be installed in multiple ways:

  1. Install using npm or your OS's package manager
  2. Install with mason-lspconfig.nvim
  3. copilot.lua and copilot.vim both bundle the LSP server in their plugin
  4. Sign in to Copilot: :LspCopilotSignIn

Requirements:

  • GitHub Copilot subscription
  • copilot-language-server installed and enabled via vim.lsp.enable

Example Configuration:

require("cursortab").setup({
  provider = {
    type = "copilot",
  },
})

blink.cmp Integration

Details

This integration exposes a minimal blink source that only consumes append_chars (end-of-line ghost text). Complex diffs (multi-line edits, replacements, deletions, cursor prediction UI) still render via the native UI.

require("cursortab").setup({
  keymaps = {
    accept = false, -- Let blink manage <Tab>
  },
  blink = {
    enabled = true,
    ghost_text = false,  -- Disable native ghost text
  },
})

require("blink.cmp").setup({
  sources = {
    providers = {
      cursortab = {
        module = "cursortab.blink",
        name = "cursortab",
        async = true,
        -- Should match provider.completion_timeout in cursortab config
        timeout_ms = 5000,
        score_offset = 50, -- Higher priority among suggestions
      },
    },
  },
})

Usage

  • Tab Key: Navigate to cursor predictions or accept completions
  • Shift-Tab Key: Partially accept completions (word-by-word for inline, line-by-line for multi-line)
  • Esc Key: Reject current completions
  • The plugin automatically shows jump indicators for predicted cursor positions
  • Visual indicators appear for additions, deletions, and completions
  • Off-screen jump targets show directional arrows with distance information

Commands

  • :CursortabToggle: Toggle the plugin on/off
  • :CursortabShowLog: Show the cursortab log file in a new buffer
  • :CursortabClearLog: Clear the cursortab log file
  • :CursortabStatus: Show detailed status information about the plugin and daemon
  • :CursortabRestart: Restart the cursortab daemon process

Development

Build

To build the server component:

cd server && go build

Test

To run tests:

cd server && go test ./...

FAQ

Which provider should I use?

See the provider feature comparison table for capabilities. For the best experience:

  • If you have a consumer GPU and want to run locally, use the sweep provider with the sweep-next-edit-1.5b model for fast local inference
  • Otherwise, use the sweepapi for the best quality with the hosted sweep-next-edit-7b model
Why are completions slow?
  1. Use a smaller or more heavily quantized model (e.g., Q4 instead of Q8)
  2. Decrease provider.max_tokens to reduce output length (also limits input context)

The github provider is known to be slower than the rest.

Why are completions not working?
  1. Update to the latest version, restart Neovim and restart the daemon with :CursortabRestart
  2. Increase provider.completion_timeout (default: 5000ms) to 10000 or more if your model is slow
  3. Increase provider.max_tokens to give the model more surrounding context (trade-off: slower completions)
How do I update the plugin?

Use your Neovim plugin manager to pull the latest changes, then restart Neovim. The daemon will automatically restart if the configuration has changed. You can also run :CursortabRestart to force a restart.

Why isn't my API key or environment variable being picked up?

The plugin runs a background daemon that persists after Neovim closes. Environment variables are only loaded when the daemon starts. If you add or change an environment variable (e.g., SWEEPAPI_TOKEN in your .zshrc), run :CursortabRestart to restart the daemon with the new environment variables.

Note: If you change plugin configuration (e.g., switch providers), the daemon will automatically restart on the next setup() call.

Contributing

Contributions are welcome! Please open an issue or a pull request.

Feel free to open issues for bugs :)

License

This project is licensed under the MIT License - see the LICENSE file for details.

About

Neovim edit completion plugin

Resources

Readme

License

MIT license
Activity
Custom properties

Stars

1 star

Watchers

0 watching

Forks

0 forks
Report repository

Releases

8 tags

Packages

 
 
 

Contributors

Languages

  • Go 90.2%
  • Lua 9.8%