Murmurator AI: built-in AI models

Not every team wants to set up and manage accounts with model providers. Starting today, every Murmurator account can use built-in open-weight models instead of (or alongside) its own keys.

What's new

• Four models, zero setup. Add Kimi K3, Qwen3.8 Max, GLM-5.3 or DeepSeek V4.1 Flash from AI models → Murmurator AI with one click. Give them a name like smart or fast and use them in any llm or agent step.
• Pay per token. Usage is billed monthly on the same invoice as your subscription. See the per-token rates.
• Spend limits. Each account has a monthly limit ($100 by default) that owners can change. Steps fail with a clear message once it's reached, so there are no surprises.
• Usage you can see. The new Usage & limits page breaks spend down by model and shows every recent call, linked to its run.
• Free credit in every trial. New trials include $5 of Murmurator AI usage so the example workflows work straight away.
• The assistant is included. Building workflows by chatting with the assistant is part of your plan and never metered.

Prefer your own keys? Nothing changes: they still work exactly as before, with nothing added on top.

Read more in the AI models guide.