Accepting applications

The Speech-to-Speech Starter Grant

For startups building on speech-to-speech.

Apply now

What’s included

144,000 minutes
12,000 minutes per month.
12 months
A full year to build and launch.
Phonic-1
The most advanced speech-to-speech model for agentic tasks.
Technical support
Dedicated Slack channel and hands-on support.

Startups building on Phonic

Speech-to-speech just became ready for production.

The Phonic team

For most of their short history, voice agents have been stitched together using STT and TTS models, but so much important context is lost in this flow. Speech-to-speech models lose nothing in translation, but until recently, they made you trade intelligence for naturalness. Phonic-1 is the most advanced speech-to-speech model for agentic tasks and provides the naturalness of speech-to-speech with the intelligence you need to run complex, high-stakes interactions.

We’re excited to work closely with the startups building products on top of speech-to-speech models. If voice is an important part of your product, we’ll cover 144,000 minutes of Phonic across 12 months and provide direct access to our research and engineering team for support in a dedicated Slack channel.

In return, we’d love to hear what works and what doesn’t. We’re still early in speech-to-speech, and developers building on top of Phonic teach us more about where we should take Phonic than any benchmark.

Who is this for?

01VC-backed startups with under $10M raised
02Voice as a core part of your product
0325 people or fewer
04Live or launching to customers within six months
05New to Phonic
06Building your own products: no agencies, consultancies, or resellers

Speaking with agents should be delightful. We’d love to help you get there.

The Phonic team

San Francisco

FAQs

Questions before you apply.

What’s included in the Starter Grant?

144,000 minutes of Phonic across 12 months (12,000 minutes per month) and direct access to our research and engineering team for support in a dedicated Slack channel. For those 12 months, you’ll get access to our Pro plan, which supports up to 50 concurrent calls, 50 agents, and 30 workspace seats. On request, we’ll also sign an MSA, DPA, or BAA.

Who’s eligible?

Early-stage, VC-backed startups (<$10M raised, ≤25 people) that are live with a voice AI product or launching to customers within six months. We review every application case by case. See the full program terms.

What happens if I exceed 12,000 minutes a month?

We bill any usage past 12,000 minutes at the end of the month at the Pro plan’s rate: $0.065 a minute today. Your rate for today’s models is locked on the day you’re approved and drops if Pro’s price does. There’s no monthly minimum during the grant, but you’ll need a payment method on file.

What happens after 12 months?

Your workspace stays on the Pro plan at its regular price: $0.065 a minute today, with a $200 monthly minimum. You can switch to pay-as-you-go on Starter instead or talk to us about volume pricing as you scale.

Can I test Phonic before I apply?

Yes. You can start building here. We are currently offering $250 a month in free credits if you sign up by November 30th. Once you’re approved, the grant replaces any credits.

How quickly will we hear back?

Our team will reach out to you within 1–2 business days.

Start building with speech-to-speech.

Apply now