SpeechPath captions come faster with comparable accuracy in FCC tests

In the FCC's test lab, NexTalk's automatic SpeechPath captions arrived faster than other providers' captions. Their median word error rates of 0.8% to 7.6% were comparable to the others' 2.0% to 7.4%.

  • Not rated a person

    The FCC certified the service on lab tests, and NexTalk had sought compensation for no calls by January 2025, so no outcome for users is reported.

Depth runs from 1, a convenience or time saved, to 5, a life or a livelihood preserved.

Status
Intended benefit.
Evidence

foundstudiedreportedclaimed

An independent regulatory document supports the benefit.

Compared with

Other IP CTS providers' aggregated word error rates on the same three call scenarios. Measured.

Without the system
2.0% to 7.4%
With it
0.8% to 7.6%
The system’s part
The National Test Lab measured SpeechPath's captions against other providers' results on the same call scenarios, which isolates the software's output. The FCC does not say which captioning mode produced the other providers' results. It does not describe how SpeechPath recognizes speech or show that it learns from data.
Later
Nothing later is known. A year after the conditional grant, the FCC noted that Solen Ventures had acquired control of NexTalk and that NexTalk had not yet sought compensation for any calls.
Not yet checked
LimitsCosts and harmsLearns from data

Download data Cite Suggest a correction

Benefit type
Reduced Task Time
System
NexTalk SpeechPath ASR-only Internet Protocol Captioned Telephone Service (IP CTS)
Made by
NexTalk, Inc. (SpeechPath, "developed by and exclusively for NexTalk")
Used by
NexTalk, Inc. (NexTalk Software, Inc. after Solen Ventures acquired control in December 2024)
Where
United States
Date
, when the source was published
Kind
Aggregate result
Broad benefit
Documenting/Recording Information
Collections
Scrutiny and findings
Last revised
1 November 2026

Sources

  1. Regulatory document, by authors with no tie to the system.

    Federal Communications Commission, Consumer and Governmental Affairs Bureau. Order, DA 24-49, CG Docket No. 03-123 (conditional certification of NexTalk, Inc. to provide IP CTS on a fully automatic basis). Adopted and released 17 January 2024.

    NTL's testing generally supports NexTalk's claims. In testing by NTL, NexTalk's median Word Error Rates for three call scenarios ranged from 0.8% to 7.6%, comparable to other IP CTS providers' aggregated Word Error Rate results, which ranged from 2.0% to 7.4% for the same call scenarios.

  2. Regulatory document, by authors with no tie to the system. Context or contrary evidence.

    Federal Communications Commission, Consumer and Governmental Affairs Bureau. Public Notice, DA 25-73, CG Docket No. 03-123: Comment Sought on Application of Solen Ventures and NexTalk for Conditional Certification as a Provider of Internet Protocol Captioned Telephone Service. Released 17 January 2025.

    NexTalk, which has not yet sought compensation from the TRS Fund, states it will not do so until after grant of the Revised Application.

  3. Other, by the makers. Context or contrary evidence.

    NexTalk. NexTalk Approved by the FCC as an IP CTS Provider: Meet SpeechPath (blog.nextalk.com), 23 December 2025. Accessed 1 November 2026.

    NexTalk was officially approved by the FCC as a provider of IP CTS (Internet Protocol Captioned Telephone Service).

Details

AI system

Workflow

Automatic speech recognition captioned every call, with no communications assistant. SpeechPath ran on Windows-based, internet-connected devices. NexTalk used commercial data centers with redundancy and backup power.

Follow-up

Conditions for replication

SpeechPath may not record, store or cache the content of any call after it ends. The service must meet the FCC's minimum standards in actual operation.

Entry details

Sector

Telecommunications relay services / accessibility for people with hearing loss.

Purpose

NexTalk built SpeechPath to caption phone calls for people with hearing loss on Windows devices, with automatic speech recognition alone.

30 more questions not yet checked

model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.

Cite this record

Peter H. Vartanian. (2026). SpeechPath captions come faster with comparable accuracy in FCC tests (AIBD-1003). Artificial Intelligence Benefits Database, edition of 1 November 2026. https://petervartanian.xyz/aibd/records/AIBD-1003.html (archived at https://doi.org/10.7910/DVN/AADSII, version 7.0)

BibTeX Data Field definitions