Retell says OpenAI models power more than 100 million monthly calls
Original title: Retell uses OpenAI models to help power 100M monthly calls
Retell reports volume grew from 30 million calls early in 2026, with OpenAI generating responses on 80% of calls.
OpenAI published a customer story on Retell AI, a voice platform for contact centers that it says serves more than 4,000 businesses. Monthly volume grew from 30 million calls in early 2026 to more than 100 million by September, and Retell says it has consumed more than 1 trillion OpenAI tokens. Retell reports resolution rates above 80% and, in its October 13, 2026 benchmark, GPT-6 Astra first on instruction following at 89.7%. Named customer results include Anker at 80.4% case resolution and Pine Park Health recovering time equal to 2.1 medical-assistant roles.
Unconfirmed: Call volumes, resolution rates, benchmark scores, and revenue figures are Retell’s or OpenAI’s claims.
- Retell cites an annual revenue run rate of $1.6 million per employee.
- Pre-launch evals had GPT-5.5 at 88.0% non-hallucination and 88.3% tool-call correctness.
- GPT-6 Astra and GPT-6 Sol median first-sentence latency was 883 milliseconds.
- GPT-5.6 Luna ranked highest on that latency measure at 497 milliseconds.
- Models named in production include GPT-6.1 Sol, GPT-6 Astra, GPT-5.5, and GPT-5.6 Terra.
- Vendors
- OpenAI
- Products
- CodexRealtime API
- Organizations
- Retell AIAnkerPine Park Health
From the original article
7 min read at openai.comGPT-6 Astra tops Retell’s voice agent benchmark, while Codex helps it reach a $1. 6M annual revenue run rate per employee.
openai.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.