Retell, providers of an artificial intelligence voice agent platform for contact centers, has launched Tell-1, a language model built specifically for live contact center calls.
Trained on hundreds of millions of calls, Tell-1 is designed for when a caller gives half an answer, changes the subject mid-task, or needs the agent to look something up and act, all while someone waits on the line.
Retell built Tell-1 to deliver the lowest possible latency and cost without sacrificing call quality. On Retell's live benchmark, Tell-1 is currently responding in less than 100 milliseconds, and it is built to stay under 300 milliseconds even at peak call volumes. Across instruction following, grounding, tool calling, colloquial quality, and latency, Tell-1 posted the highest overall score of any model tested.
"Enterprises need to be able to trust voice AI agents with the calls that matter," said Bing Wu, co-founder and CEO of Retell, in a statement. "For AI voice agents, the bottleneck isn’t intelligence but timing: the model has to reason through the business’s rules and act in the time it takes a person to pause, and that’s the problem we built Tell-1 to solve."
Tell-1 uses an interleaved reasoning architecture to allow the model to think while speaking. Rather than treating each turn as an isolated question, the model reasons through the conversation as it unfolds: which of the business's rules apply now, which do not, which information is still missing, and whether to ask, respond, or act. In practice, Tell-1 helps AI voice agents do the following:
- Follow long, complex instructions: Keeping business goals, policies, and exceptions in view from the first minute of a call to the last.
- Work with partial answers: When a caller gives only a first name, the agent asks for the missing detail instead of guessing or restarting the form.
- Handle detours: The agent acknowledges an off-topic question, then returns to the task without sounding like the conversation has reset.
- Act at the right time: Calling the right tool, such as an availability or account lookup, once it has enough information and using the result to move the call forward.
Retell also introduced Retell Dialer for voice agents to orchestrate outbound campaigns across AI voice and SMS agents. Where Tell-1 handles decisions during a live conversation, Dialer turns the outcome of that conversation into the next action. A requested callback replaces the next planned attempt and resumes the same conversation at the contact's requested local time; a qualified prospect moves to a specialist agent; and someone interested but not ready receives an SMS follow-up from a nurture agent. Retell Dialer also brings together contact screening, local presence and calling windows, number rotation and spam monitoring, configurable retries, and CRM integration.