Cheapest real-time / streaming transcription API
Every directly-comparable API here supports streaming, so the cheapest real-time choice is simply the cheapest overall: AssemblyAI Universal-3 Pro at $3.50/1000 min, with roughly 150 ms latency. ElevenLabs Scribe v2 is a close second on price; Deepgram Nova-3 bills per second, which can suit short or bursty live workloads. Token-priced LLM models are excluded as their per-minute cost isn't comparable. A light estimate from public pricing.
DefaultAssemblyAI Universal-3 ProgoedkoopsteAssemblyAI Universal-3 Pro
Provider offerings
| Offering | Price ($/1000 min) | WER | Langs | Latency | Capabilities |
|---|---|---|---|---|---|
| ElevenLabs Scribe v2ElevenLabs | 3.67 | 2.2% | 90 | 150 ms | diarizationtimestampsvocab |
| AssemblyAI Universal-3 ProAssemblyAI | 3.5 | 3.1% | 99 | 150 ms | diarizationtimestampsvocab |
| Deepgram Nova-3Deepgram | 4.3 | 5.2% | 50 | 300 ms | diarizationtimestampsvocab |
| Speechmatics EnhancedSpeechmatics | 6.7 | 4% | 70 | 500 ms | diarizationtimestampsvocab |
| OpenAI gpt-4o-transcribeOpenAI | 6 | 4% | — | — | — |
| Google Gemini 3 FlashGoogle | 1.92* | 2.9% | — | — | — |
* token-/credit-priced — the headline understates real per-unit cost, so it is excluded from the cheapest ranking.
Sources
- Artificial Analysis — Speech to Text2026-06-19
- Open ASR Leaderboard2026-06-19