How to use the http API POST chat/completions

[Dronakurl]

After setting up tabby with the docker compose file in quick start, I wanted to look into implementing a chat completion for my favourite editor neovim.

I wanted to use codecompanion.nvim for this, as this plugin allows to specify new adapters easily.

However, I got stuck, figuring out the http API to just send basic chat completion requests.
The API documentation doesn't help and I couldn't find the documentation of the payload for the chat/completions endpoint. completions works without problems.

Can someone please help me make this first step?

POST http://localhost:8080/v1/chat/completions HTTP/1.1
accept: application/json
content-type: application/json
Authorization: Bearer auth_881a958cbb454e1aa87a3fa17e8ex649

{
  "messages": ["Something"]
}

--> Failed to deserialize the JSON body into the target type: messages[0]: data did not match any variant of untagged enum ChatCompletionRequestMessage at line 2 column 23

[wsxiaoys]

The chat completions api is an openai like interface, could you try following curl command (with your bearer token):

curl http://localhost:8080/v1/chat/completions -X POST --data-raw '{"messages": [{"role": "user", "content": "hi"}]}'

[Dronakurl]

Great, thank you, I managed to configure codecompanion.nvim for usage with chat using this configuration:

{
	"olimorris/codecompanion.nvim",
	dependencies = {
		"nvim-lua/plenary.nvim",
		"nvim-treesitter/nvim-treesitter",
	},
	config = function()
		require("codecompanion").setup({
			strategies = {
				chat = { adapter = "tabby" },
				inline = { adapter = "tabby" },
			},
			adapters = {
				tabby = function()
					return require("codecompanion.adapters").extend("openai_compatible", {
						env = {
							url = "http://localhost:8080", -- optional: default value is ollama url http://127.0.0.1:11434
							api_key = "auth_881a958cbb454e1aa87a3fa17e8e9649", -- optional: if your endpoint is authenticated
							chat_url = "/v1/chat/completions", -- optional: default value, override if different
						},
						schema = {
							model = {
								default = "Qwen2-1.5B-Instruct",
								choices = { "Qwen2-1.5B-Instruct" },
							},
						},
					})
				end,
			},
		})
	end,
}

One problem was, that tabby doesn't provide a GET /v1/models endpoint. That's why, you have to explicitely specify the chat model, you use