# Phonic Speech-to-speech voice agent infrastructure for teams building phone and audio workflows. - Findings24: - ウェブサイト: - 料金: Freemium - カテゴリー: AIエージェント基盤, AIオーディオAPI ## Phonicとは？ Phonic is a speech-to-speech voice agent platform for teams building conversational phone and audio workflows. Its audio models process input speech directly to produce natural spoken responses, while frontier models support reasoning and tool calling. The platform is designed to fit into existing systems through options including LiveKit, webhooks, WebSockets, and no-code integrations. Example workflows include recruiting screens, logistics dispatch calls, patient intake, and customer support. ## Phonicでできること ### Audio-conditioned conversation Agents condition responses on input audio, including cues such as tone, hesitation, and pronunciation, rather than relying only on a text transcript. ### Speech-to-speech responses The platform takes speech as input and returns spoken output, with the site describing response latency as under a second. ### Tool calling and reasoning Frontier models supervise agent reasoning and tool calling, supporting workflows that need to act on information gathered in a call. ### Integration options Teams can connect agents through LiveKit, webhooks, WebSockets, or more than 1,000 no-code integrations. ### Multilingual calls Phonic supports more than 50 languages, allowing agents to conduct conversations in callers' languages. ### Deployment and enterprise controls Options include multi-tenant Phonic Cloud, dedicated infrastructure, VPC peering, and enterprise self-hosting. The site also lists SOC 2 Type II certification, SSO, and HIPAA business associate agreements. ## 利用シーン ### Recruiting phone screens Recruiting platforms can use agents to ask open-ended screening questions, capture candidate responses, and connect the workflow to an ATS. ### Logistics dispatch and driver calls Logistics teams can automate check-ins and dispatch conversations involving pickup and drop-off details, load identifiers, weights, and rates. ### Patient intake Healthcare teams can use voice agents to gather information from patients through conversational intake calls. ### Customer support Support organizations can use agents to handle customer calls as part of a voice workflow, with spoken interaction rather than a text-only interface. ## よくある質問 ### Which languages can agents speak? The site states that Phonic supports more than 50 languages and describes agents switching to languages callers are comfortable using. ### How can a team connect Phonic to its existing stack? The listed options include LiveKit, webhooks, WebSockets, and no-code integrations. The recruiting page also mentions a no-code builder and SDK. ### What deployment options are available? Listed options include default multi-tenant Phonic Cloud, dedicated infrastructure in a region of choice, VPC peering, and enterprise self-hosting. ### How is usage priced? The pricing page lists a per-minute conversation charge, with free credits on Starter and Pro plans, alongside custom enterprise pricing. The listed per-minute rates are $0.14 on Starter and $0.12 on Pro; the page also lists Pro at $240 per month. ## クイック情報 - **Product category**: Speech-to-speech voice agent platform - **Primary users**: Teams building recruiting, logistics, healthcare, and support voice workflows - **Language support**: 50+ languages - **Integration methods**: LiveKit, webhooks, WebSockets, and no-code integrations - **Pricing model**: Per-minute conversation pricing; custom enterprise pricing is listed - **Website**: phonic.ai