no longer getting the model choices picker

[atkr]

Recently came back from a month long vacation and I noticed that I am no longer getting a picker to choose from the different choices of models.

I am using multiple adapters (openai_compatible, ollama and a custom open-webui adapter). When I hit ga from the chat window, after hiting enter to select an adapter, I used to be prompted to select a model. I no longer get that!

Using ollama as an example, I have this adapter config:

      locallama_ctx40_qwen_think = function()
        return require('codecompanion.adapters').extend('ollama', {
          formatted_name = 'locallama',
          schema = {
            model = { default = 'hf.co/unsloth/Qwen3-32B-GGUF:UD-Q6_K_XL' },
            num_ctx = { default = 40960 },
            temperature = { default = 0.6 },
            top_k = { default = 40 },
            top_p = { default = 0.95 },
          },
        })
      end,

After opening a new chat window and selecting this adapter, I see this in the debug window (gd):

-- Adapter: "locallama"
-- Buffer: 8
-- Context: "codecompanion.lua" (1)

local settings = {
  model = "hf.co/unsloth/Qwen3-32B-GGUF:UD-Q6_K_XL",  -- "hf.co/unsloth/QwQ-32B-GGUF:UD-Q6_K_XL", "hf.co/unsloth/Qwen3-32B-GGUF:UD-Q5_K_XL", "hf.co/unsloth/QwQ-32B-GGUF:UD-Q5_K_XL", "hf.co/unsloth/Qwen3-30B-A3B-GGUF:UD-Q6_K_XL", "hf.co/unsloth/Qwen3-14B-GGUF:UD-Q6_K_XL", "hf.co/unsloth/GLM-4-32B-0414-GGUF:Q5_K_XL", "hf.co/unsloth/GLM-4-32B-0414-GGUF:Q6_K_XL", "hf.co/unsloth/Qwen2.5-Coder-32B-Instruct-GGUF:Q5_K_M", "hf.co/unsloth/Qwen2.5-Coder-32B-Instruct-GGUF:Q6_K", "hf.co/unsloth/Mistral-Small-3.1-24B-Instruct-2503-GGUF:UD-Q6_K_XL", "hf.co/unsloth/Qwen2.5-Coder-14B-Instruct-GGUF:Q6_K", "hf.co/unsloth/phi-4-GGUF:Q5_K_M", "hf.co/darkc0de/gemma-3-27b-it-abliterated-Q5_K_M-GGUF:latest", "hf.co/unsloth/gemma-3-27b-it-GGUF:UD-Q6_K_XL", "hf.co/unsloth/gemma-3-27b-it-GGUF:UD-Q5_K_XL", 
  temperature = 0.6,
  num_ctx = 40960,
  mirostat = 0,
  mirostat_eta = 0.1,
  mirostat_tau = 5,
  repeat_last_n = 64,
  repeat_penalty = 1.1,
  seed = 0,
  stop = nil,
  num_predict = -1,
  top_k = 40,
  top_p = 0.95,
}

local messages = 
{ {
    content = "You are an AI programming assistant named \"CodeCompanion\". You are currently plugged in to the Neovim text editor on a user's machine.\n\nYour core tasks include:\n- Answering general programming questions.\n- Explaining how the code in a Neovim buffer works.\n- Reviewing the selected code in a Neovim buffer.\n- Generating unit tests for the selected code.\n- Proposing fixes for problems in the selected code.\n- Scaffolding code for a new workspace.\n- Finding relevant code to the user's query.\n- Proposing fixes for test failures.\n- Answering questions about Neovim.\n- Running tools.\n\nYou must:\n- Follow the user's requirements carefully and to the letter.\n- Keep your answers short and impersonal, especially if the user responds with context outside of your tasks.\n- Minimize other prose.\n- Use Markdown formatting in your answers.\n- Include the programming language name at the start of the Markdown code blocks, do not forget triple backticks around code blocks.\n- Avoid including line numbers in code blocks.\n- Avoid wrapping the whole response in triple backticks.\n- Only return code that's relevant to the task at hand. You may not need to return all of the code that the user has shared.\n- Use actual line breaks instead of '\\n' in your response to begin new lines.\n- Use '\\n' only when you want a literal backslash followed by a character 'n'.\n\nWhen given a task:\n1. Think step-by-step and describe your plan for what to build in pseudocode, written out in great detail, unless asked not to do so.\n2. Output the code in a single code block, being careful to only return relevant code.\n3. You should always generate short suggestions for the next user turns that are relevant to the conversation.\n4. You can only give one reply for each conversation turn.\n",
    cycle = 1,
    id = 967016408,
    opts = {
      tag = "from_config",
      visible = false
    },
    role = "system"
  } }

where we can see in the comment on the settings.model line that all other available choices are indeed getting pulled from ollama.. but I'm not longer getting prompted to choose from a picker view. Why is that?

I have tried:

• not selecting a default model in the adapter config -> same behavior
• I have enabled TRACE logging, nothing at all is logged in regards to model choices.

Any help will be greatly appreciated :)
edit: some typos

[ravitemer]

The code for that is at codecompanion.nvim/lua/codecompanion/strategies/chat/keymaps.lua. It uses vim.ui.select which might be overridden by some other plugin. You can add some vim.notify statements in the following function in the file to see what's going wrong.

M.change_adapter = {
  desc = "Change the adapter",
  callback = function(chat)
    if config.display.chat.show_settings then
      return util.notify("Adapter can't be changed when `display.chat.show_settings = true`", vim.log.levels.WARN)
    end

    local function select_opts(prompt, conditional)
      return {
        prompt = prompt,
        kind = "codecompanion.nvim",
        format_item = function(item)
          if conditional == item then
            return "* " .. item
          end
          return "  " .. item
        end,
      }
    end

    local adapters = vim.deepcopy(config.adapters)
    local current_adapter = chat.adapter.name
    local current_model = vim.deepcopy(chat.adapter.schema.model.default)

    local adapters_list = vim
      .iter(adapters)
      :filter(function(adapter)
        return adapter ~= "opts" and adapter ~= "non_llms" and adapter ~= current_adapter
      end)
      :map(function(adapter, _)
        return adapter
      end)
      :totable()

    table.sort(adapters_list)
    table.insert(adapters_list, 1, current_adapter)

    vim.ui.select(adapters_list, select_opts("Select Adapter", current_adapter), function(selected)
      if not selected then
        return
      end

      if current_adapter ~= selected then
        chat.adapter = require("codecompanion.adapters").resolve(adapters[selected])
        util.fire(
          "ChatAdapter",
          { bufnr = chat.bufnr, adapter = require("codecompanion.adapters").make_safe(chat.adapter) }
        )
        chat.ui.adapter = chat.adapter
        chat:apply_settings()
      end

      -- Update the system prompt
      local system_prompt = config.opts.system_prompt
      if type(system_prompt) == "function" then
        if chat.messages[1] and chat.messages[1].role == "system" then
          local opts = { adapter = chat.adapter, language = config.opts.language }
          chat.messages[1].content = system_prompt(opts)
        end
      end

      -- Select a model
      local models = chat.adapter.schema.model.choices
      if not config.adapters.opts.show_model_choices then
        models = { chat.adapter.schema.model.default }
      end
      if type(models) == "function" then
        models = models(chat.adapter)
      end
      if not models or vim.tbl_count(models) < 2 then
        return
      end

      local new_model = chat.adapter.schema.model.default
      if type(new_model) == "function" then
        new_model = new_model(chat.adapter)
      end

      models = vim
        .iter(models)
        :map(function(model, value)
          if type(model) == "string" then
            return model
          else
            return value -- This is for the table entry case
          end
        end)
        :filter(function(model)
          return model ~= new_model
        end)
        :totable()
      table.insert(models, 1, new_model)

      vim.ui.select(models, select_opts("Select Model", new_model), function(selected)
        if not selected then
          return
        end

        if current_model ~= selected then
          util.fire("ChatModel", { bufnr = chat.bufnr, model = selected })
        end

        chat:apply_model(selected)
        chat:apply_settings()
      end)
    end)
  end,
}

[atkr]

Thanks for the suggestion! I will try troubleshooting as you suggested. Will report back with findings.

[atkr]

I found the change the led to this behavior change:

90e82ab Pedro Ferrari (2025-05-04 16:47):
feat(adapters): can automatically choose default model when switching adapters (#1369)

Hunk 1 of 1 @@ -509 +510,3 @@
+      if not config.adapters.opts.show_model_choices then
+        models = { chat.adapter.schema.model.default }
+      end

I modified my config to include the config to always show model choices:

  opts = {
    adapters = {
      opts = {
        show_model_choices = true, -- <<< ADDED THIS
      },
    },
  }

[zbindenren]

thanks, that helped me