Syl
Intro
Syl is the flagship short-utterance speech-recognition (ASR) model Bolna named in public on 20 November 2026. The official account @bolna_dev posted Introducing Syl at 2026-11-20T12:30:52Z (20:30 Beijing time): “Introducing Syl, Bolna’s flagship short-utterance model that catches the smallest of affirmations and transcribes the way India actually speaks… not an easy job, but it’s what we signed up for! More details to follow soon.”
The same day, Bolna held Symphony 2026 at The Lalit Ashok in Bengaluru. The official agenda lists a “Launch / Bolna Launches” session on the main stage (Sur) at 16:45–17:30 IST, hosted by Siddharth Batra and Prabal Rastogi. Two days earlier, the 16 November newsroom post Bolna brings India’s Voice AI ecosystem together at Symphony 2026 in Bengaluru had already said Bolna would launch in-house speech models at Symphony, starting with an ASR model built for the short utterances common in customer calls.
This entry is for the officially named model Syl. The official post says more details will follow, so this page does not treat an unpublished model card, pricing, or language list as fact. The product site is bolna.ai. It is not the Bolna platform itself, and it is not another vendor’s speech-recognition model.
Advertisement
Explain
In the official wording, Syl is a short-utterance ASR model: it is meant to catch brief affirmations and transcribe speech the way people in India actually talk, rather than only in formal or single-accent speech. That matches the 16 November newsroom preview of an in-house ASR model built for the short utterances common in customer calls. Bolna describes itself as a voice AI platform built for India; the same newsroom post notes that calls there often switch languages, include regional accents, and come from noisy environments. That is context for the platform, not a published benchmark for Syl.
At the time of this entry, Bolna had not yet published a standalone model page, weights, API name, or language list. The company announced the name on X and said more details would follow. This page therefore does not treat third-party claims about parameter count, word-error rate, open weights, or product packaging as fact, and it does not fold other Symphony announcements into this name.
On the timeline: the newsroom preview ran on 16 November; the public naming was at 12:30 UTC on 20 November, shortly after the scheduled end of the “Bolna Launches” session (17:30 IST is 12:00 UTC). This entry follows the official post, the agenda, and the newsroom article.
Usage
For now, start from the official post Introducing Syl, the Symphony agenda, and the 16 November newsroom article Symphony 2026 preview. If you want to use Bolna’s in-house recognition model in a voice agent, wait for the follow-up docs, console, or API notes the company said would come later. Until then, do not assume a public model ID or calling convention.
The stated use is short utterances and brief confirmations typical of customer calls, not general long-form or meeting transcription. Official details on pricing, language coverage, open weights, and availability had not been published with the name. This page only restates those official sources and does not treat Syl as another name for the Bolna platform.