Models in Workspace
The models you can choose
Mthandizi FastOpenAI · Included
Fast everyday answers, and the model that runs career tools
GPT-5.6 LunaOpenAI · Efficient
Fast, economical everyday work
Claude Haiku 5.5Anthropic · Efficient
Fast structured reading and drafting
GPT-5.6 TerraOpenAI · Advanced
Balanced professional reasoning
Claude Sonnet 5.5Anthropic · Advanced
Complex analysis and polished drafting
GPT-5.6 SolOpenAI · Frontier
Frontier capability for difficult professional work
Claude Opus 5.5Anthropic · Frontier
Premium reasoning for demanding knowledge work
A more capable model uses more credits for the same answer. The picker shows which models your plan offers.
Which plans offer which models
The Free plan answers with Mthandizi Fast. Paid plans can choose any model in the list above. See the plans for what each one includes.
Switching model inside a chat
On a paid plan you can change model in the middle of a chat: pick another one in the model picker, and your next message is answered by it. When you open the chat again, a line marks where it switched.
Answers come from the company that makes the model: OpenAI or Anthropic. When you switch between them, the conversation so far goes to the new one with your next message, through the same privacy protections. The privacy policy explains them.
A few things stay on one model
- A Mthandizi Code web project keeps the model it started with, because its plans and prices were made for that model.
- Career tools (matching your CV, tailoring it, cover letters) run on Mthandizi Fast, the model wired to them, whichever model is selected; a draft can then be rewritten with the model you chose.
- A chat that is already on an earlier Claude model keeps it until you switch; earlier models are not offered for new chats.
Using the coding agent in your terminal? Its models are chosen with /model: see Mthandizi Code.