Smallest AI
Also known as: Smallest.ai, Atoms, Waves, Lightning, Pulse
Full stack voice AI platform combining Atoms, a no code voice and chat agent builder handling conversations in more than twenty languages, with its own co optimized models: Lightning text to speech at roughly one hundred millisecond latency and Pulse speech to text at sixty four millisecond first transcript.
Smallest AI was founded by Sudarshan Kamath and Akshat Mandloi and has raised eight million dollars in seed funding led by Sierra Ventures, months after a pre seed led by 3one4 Capital, with the batch Crunchbase data listing San Francisco as headquarters and a non equity assistance event in October 2025. The company's thesis is precision over scale: many small, efficient models that each know exactly what matters, co optimized to run together rather than assembled from different vendors. In its first year it cut text to speech latency from roughly two seconds to one hundred milliseconds, dropped TTS cost from twenty cents to a penny per minute at scale, and moved from prototype to powering millions of enterprise calls each month.
The stack has three layers. Lightning is the text to speech line, with V3 hitting about one hundred millisecond latency across fifteen plus languages including Indic and European coverage, instant voice cloning from ten seconds of audio, and streaming over HTTP, SSE, or WebSocket. Pulse is streaming speech to text with a sixty four millisecond time to first transcript. Atoms is the agent platform on top: no code Single Prompt and Conversational Flow agents with an Agentic Graph Builder, knowledge bases, webhooks, API calls from within conversations, phone numbers, outbound campaigns and audiences, a web voice and chat widget, post call metrics, conversation logs, versioning, and testing, all documented alongside Node.js and Python SDKs, a WebSocket SDK, mobile integrations, and an MCP section in the developer docs. The full agent pipeline stays under eight hundred milliseconds per turn, and the whole stack is available on AWS Marketplace under consolidated billing.
Smallest AI fits teams building production voice agents for support, sales, and operations who want the model layer and agent layer from one vendor with published usage pricing, and developers who need fast multilingual TTS and STT as raw APIs. It competes with Vapi, Retell, ElevenLabs, Cartesia, and Deepgram across its layers.
Vendor details
Canonical URL
https://smallest.ai
Category
Voice agent
Subcategory
Full stack voice AI: Atoms agents plus Lightning TTS and Pulse STT
Funding status
Independent, founded by Sudarshan Kamath and Akshat Mandloi. Raised eight million dollars in seed funding led by Sierra Ventures, months after a pre seed led by 3one4 Capital, and reports powering millions of enterprise calls per month within a year of launch. The batch Crunchbase data lists San Francisco headquarters and a non equity assistance event in October 2025.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
Atoms agents connect to business tools and data, make API calls mid conversation, and fire webhooks, with an integrations catalog, phone number provisioning, and an embeddable web voice and chat widget. Developers get Node.js and Python SDKs, a WebSocket SDK, mobile integrations, REST APIs across Waves and Atoms, an MCP section in the docs, and llms.txt indexes for AI agents. The full stack is procurable through AWS Marketplace with consolidated billing.
In practice
A support team ships a phone agent in an afternoon: four short prompts generate a Single Prompt agent with voice, model, and knowledge base configured, tested over a web call before going live on a real number.
A developer building a voice assistant needs speech that keeps up with conversation. Lightning V3 returns audio in roughly one hundred milliseconds and Pulse transcribes with a sixty four millisecond first token, keeping each turn under eight hundred milliseconds.
An operations team runs outbound campaigns in twenty plus languages, with agents collecting information, completing transactions, and triggering downstream actions through API calls, while post call metrics and conversation logs feed review.
Sources & related URLs
Related / legacy domains
Research sources
Capability coverage
9.0 / 14 capabilities · 64%
| Integrations & Tool CallingAtoms use business context, tools, and data, make API calls from within conversations, trigger downstream actions, and fire webhooks, with an integrations catalog documented, AWS Marketplace listing and Atoms docs retrieved 2026-07-08 | Full |
|---|---|
| Workflow OrchestrationConversational Flow agents with a workflow tab handle multi step conversations, and all plans include the Agentic Graph Builder for orchestrated flows, with end to end conversation handling documented, Atoms docs and billing page retrieved 2026-07-08 | Full |
| Knowledge Grounding & RAGKnowledge Base is a first class documented feature configured per agent, grounding conversations in business content, Atoms docs navigation and agent creation guide retrieved 2026-07-08 | Full |
| Human Oversight & GuardrailsNo human handoff, escalation, or approval workflow is documented on retrieved pages, Atoms docs retrieved 2026-07-08 | Unable to verify |
| Security, Identity & GovernanceOfficial pages describe secure, compliant, production ready performance with a support and security docs section, while a third party profile lists SOC 2 Type 2, HIPAA, PCI, GDPR, and ISO aligned standards not confirmed on retrieved official pages, smallest.ai and voiceaispace profile retrieved 2026-07-08 | Partial |
| Observability & AuditabilityAnalytics and logs are first class documented features including conversation logs, post call metrics, and analytics dashboards, Atoms docs navigation retrieved 2026-07-08 | Full |
| Memory & State PersistenceAtoms continuously learn from mistakes and the company positions models that learn continuously from interaction with access to external memory, though a documented per customer memory feature was not retrieved, AWS Marketplace listing and smallest.ai homepage retrieved 2026-07-08 | Partial |
| Deployment & Data ResidencyThe platform is delivered as cloud SaaS; edge deployment references concern the company's small model philosophy rather than customer self hosting of the Atoms platform, and no residency options are documented, smallest.ai and Atoms docs retrieved 2026-07-08 | Unable to verify |
| Prebuilt Agents, Templates & PacksWorkflow templates ship for customer support, appointment scheduling, sales, and lead qualification, and Create with AI generates a complete Single Prompt agent from four short prompts, company blog and Atoms docs retrieved 2026-07-08 | Full |
| Triggers & Channel CoverageCoverage spans inbound and outbound phone via provisioned numbers, campaigns and audiences, a web voice and chat widget, chat and email and social channels, and webhooks, Atoms docs and third party platform profile retrieved 2026-07-08 | Full |
| Model Flexibility & RoutingAgent configuration includes a model setting alongside voice and knowledge base, but a documented list of selectable LLM providers was not retrieved, Atoms docs agent creation guide retrieved 2026-07-08 | Partial |
| APIs, SDKs & MCP ExtensibilityREST APIs across Waves and Atoms, Node.js and Python SDKs, a WebSocket SDK, mobile integrations, an MCP section in the developer docs, and llms.txt indexes for AI agents, smallest.ai homepage, Atoms docs, and third party profile retrieved 2026-07-08 | Full |
| Testing, Debugging & OptimizationA testing section is documented with web call test runs before deployment and post call metrics for production review, though a scenario based evaluation suite is not documented, Atoms docs retrieved 2026-07-08 | Partial |
| Browser & Computer UseNo browser or computer use capability is described on retrieved pages, smallest.ai and Atoms docs retrieved 2026-07-08 | Unable to verify |
Pricing
Atoms voice agent usage starts at eight cents per minute per the company FAQ, dropping two to three times lower at volume, with a free tier to start building
per minute usage for voice agents plus subscription plans on monthly or annual billing with credits tracked in the dashboard
What is public
The Atoms FAQ states usage starts at eight cents per minute and declines two to three times at volume. The pricing page offers a free start and per minute rates for Pulse speech to text, and the billing docs describe monthly or annual subscription plans with included credits, auto purchase, and self serve plan management. Exact plan tiers and per credit rates were not fully retrievable.
Billing mechanics
Self serve subscription management at app.smallest.ai with monthly or annual billing, credit tracking, and plan upgrades or downgrades in the dashboard. All plans include the no code Agent Builder and Agentic Graph Builder. The full stack is also procurable through AWS Marketplace with consolidated billing. Enterprise developer support applies at one hundred thousand dollars or more in annual account size.
Cost watchouts
Telephony carries per minute economics on both agent usage and phone numbers, and long average call durations multiply spend faster than call counts suggest.
Variable cost rationale
Cost tracks conversation minutes: the official FAQ puts Atoms at eight cents per minute falling two to three times lower with volume, so spend scales directly with call volume and duration. Subscription plans layer monthly credits on top, and the company markets TTS at a penny per minute at scale, so the usage axis dominates.
Additional watchouts
Volume discounts are real but unspecified: the FAQ says rates go two to three times lower as volumes scale, without publishing the thresholds, so negotiate the curve rather than accepting the entry rate at scale.
Sales call required
Mixed (some tiers require a call)
Free / trial
Free tier to start building voice agents, per the official pricing page
Key ambiguities
Where the volume thresholds sit between the eight cent entry rate and the two to three times lower rates at scale.
Missing data
Full plan tier pricing, per credit rates, and volume discount thresholds were not retrievable from fetched pages.
Related vendors
- aiola — Enterprise voice AI for field and frontline teams: the Jargonic ASR…
- Broccoli — AI voice agents built for the trades, answering every call for…
- Consio — AI phone platform for Shopify brands: a twenty four seven inbound AI…
- Equal AI — AI call assistant that answers, screens, and manages incoming phone…
- Giga — Enterprise voice and chat AI agents that resolve customer support…
- Gnani.ai — Voice AI platform with foundational speech models and agents that…