S
Speko
The router for voice models: every speech model, benchmarked language by language, wired into one API.
Backed by Y Combinator21 speech models benchmarkedProvider-neutral STT · LLM · TTSLiveKit + Pipecat native gatewayLanguage-level benchmarks across 9 languages
Opportunity score
72/100
Speko has a strong wedge in language-specific voice model routing with developer-friendly distribution, but the business remains unproven on revenue, retention and defensibility.
Founder verdict
MAYBE
Speko is a strategically positioned YC-backed voice routing layer with a data-led wedge, but the moat is still unproven at scale.
What is Speko?
Speko is a YC-backed voice AI infrastructure startup that positions itself as a provider-neutral router for speech-to-text, LLM, text-to-speech, and speech-to-speech workloads. Its core wedge is a language-by-language benchmark table showing that no single model wins everywhere, which it uses to justify a managed routing API. It meets developers where they already build by offering native LiveKit and Pipecat gateway integrations, BYOK credentials, OpenAPI/AsyncAPI typed contracts, and MCP support. The business model and pricing are not publicly disclosed, but the product strategy is clearly aimed at becoming the data-backed control plane for voice agent stacks.
Get the full Speko playbook
The complete reverse-engineering — what to copy, what to avoid, and exactly what to build instead.
- Full opportunity score across 8 dimensions
- The real problem & why customers pay
- Why it's winning — with evidence
- Complete business reverse-engineering
- Competitive advantages & moat analysis
- Weaknesses, risks & what to avoid
- Clone strategy — what to build instead
- Week-by-week MVP roadmap
- Recommended technical stack
- Founder verdict & highest-leverage move