Soniox Speech-to-Text
Budget speech-to-text that punches above its price, with translation built in
About
Soniox is a speech-to-text API that competes on price and multilingual chops rather than brand. It offers real-time WebSocket transcription ($0.12/hr) and async REST ($0.10/hr) — a fraction of OpenAI or Azure rates — with 60+ languages, any-to-any translation, mid-sentence code-switching, custom vocabulary, and speaker diarization in a single call. Vendor benchmarks claim 6.5% English WER; a third-party 2026 streaming benchmark measured v5 Real-Time at 4.5% final WER with 0.05s final latency, inside the accuracy top 10. Community testers praise its multilingual accuracy and a generous $200 no-card API credit. Caveats: vendor accuracy claims are self-reported, it is a smaller company than the hyperscaler alternatives, and enterprise compliance buyers should verify certifications against their own requirements (SOC 2, HIPAA, GDPR claimed).
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Vendor-reported 6.5% English WER; third-party 2026 benchmark: 4.5% final WER at 0.05s latency
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Vendor benchmarks claim 6.5% English WER; third-party 2026 benchmark measured 4.5% final WER; Pricing $0.10-0.12/hr and $200 credit corroborated across community research docs; Smaller vendor; compliance claims should be verified by buyer.