Resolve model currently in use in chat

[ruicsh]

I'm trying to display the current model in use on the statusline. I can't use the config, because different prompts will use different models. Example: the default model is claude for dev prompts, but then I have o1 for generic chats. Is there a public API exposed for this?

[deathbeam]

Nope, but I plan to rework how models/agents/contexts are stored (e.g instead of storing the value passed in ask on chat config, store it as sticky prompt, which is something that you can see and also parse easily)

[deathbeam]

Should be possible to resolve easily now after: #855

Through: chat.resolve_model

Just be aware this is using plenary.async so when using it ideally use it inside of async.run(function()
end) call

[ruicsh]

works great, thanks!

[ouuan]

This works for me:

local async = require 'plenary.async'
local chat = require 'CopilotChat'
local model = ''
local resolving_model = false

local function copilot_chat_model()
  if resolving_model then
    return model
  end
  resolving_model = true
  async.run(function()
    local resolved_model = chat.resolve_model()
    if resolved_model then
      model = resolved_model
    end
    resolving_model = false
  end)
  return model
end

Without the resolving_model guard, it calls the API ~20 times (depending on network latency) before displaying the model.