llms.txt
+12 −6
− - [Speech-to-Text APIs in 2026: Pricing and Features](https://verli.app/blog/choosing-a-speech-to-text-api): Compare 17 speech-to-text APIs and self-hosted models on pricing, streaming, translation, languages, diarization, rate limits, and deployment costs.− - [6 Live Meeting Translation Tools Compared (2026)](https://verli.app/blog/best-real-time-meeting-translation-tools): Compare six live meeting translation tools on bots, captions, spoken translation, and cost, plus the built-in options in Zoom, Teams, and Google Meet.− - [Speech-to-Text Accuracy Benchmark: Engine Results](https://verli.app/blog/speech-to-text-accuracy-benchmark): Speech-to-text engines tested on identical LibriSpeech, Earnings-22, and SPGISpeech audio with one scorer.− - [Streaming Speech-to-Text: Accuracy and Latency](https://verli.app/blog/streaming-speech-to-text-benchmark): Ten speech-to-text engines tested on identical audio: streaming versus batch word error rates, first-word latency, finalization latency, and reliability.− - [Speech-to-Text Accuracy by Language: 16 Engines](https://verli.app/blog/per-language-speech-to-text-accuracy): Compare 16 speech-to-text engines across 10 languages using the same audio, word error rate, character error rate, and paired-bootstrap tests.− - [Text-to-Speech APIs Compared: Pricing and Features](https://verli.app/blog/choosing-a-text-to-speech-api): Compare 18 text-to-speech APIs on pricing, streaming, voice cloning, languages, rate limits, training-data policies, latency, and quality.+ - [Speech-to-Text APIs in 2026: Pricing and Features](https://verli.app/blog/choosing-a-speech-to-text-api): 17 speech-to-text APIs compared: Groq's hosted Whisper is cheapest at $0.04 an hour, and only Soniox, Gladia and Speechmatics translate in the live stream.+ - [8 Live Meeting Translation Tools Compared (2026)](https://verli.app/blog/best-real-time-meeting-translation-tools): Verli ($7.99 a month) and Transync ($8.99) are the cheapest bot-free tools that speak the translation. We compare eight, plus what Zoom, Teams, and Meet offer.+ - [Speech-to-Text Accuracy Benchmark: Engine Results](https://verli.app/blog/speech-to-text-accuracy-benchmark): No engine wins every test set: OpenAI ties for the lowest error on clean speech but finishes last on earnings calls, where xAI's Grok leads at 9.53%.+ - [Streaming Speech-to-Text: Accuracy and Latency](https://verli.app/blog/streaming-speech-to-text-benchmark): Streaming cost every engine but one some accuracy on clean speech and helped none on hard audio. First words took 0.90s to 6.56s to appear.+ - [Speech-to-Text Accuracy by Language: 16 Engines](https://verli.app/blog/per-language-speech-to-text-accuracy): ElevenLabs Scribe had the lowest mean error of 16 speech-to-text engines in 10 languages, but only Spanish, Korean, and Arabic had a clear winner.+ - [Text-to-Speech APIs Compared: Pricing and Features](https://verli.app/blog/choosing-a-text-to-speech-api): 18 text-to-speech APIs compared: prices span 40x, from $2.50 per million characters at Lemonfox to $100, and Cartesia was fastest in our test at 109ms.+ - [Best LLM for Translation: 11 Tested on Live Speech](https://verli.app/blog/best-llm-for-translation): GPT-6 Luna translated live speech best of 11 LLMs we tested in 24 language pairs, and DeepSeek V4.1 Flash came within a point at half the cost.+ - [Live Captions in Many Languages With One LLM Call](https://verli.app/blog/how-we-translate-live-captions): One loop per room translates a speaker's live captions into every language in one LLM call, sentence by sentence, after whole lines timed out in production.+ - [WebSocket vs SSE for Live Captions: Our Numbers](https://verli.app/blog/websocket-vs-sse): Streaming a model's answer over HTTP cut its first output by 0.4 to 1.4 s, and SSE reached 50 live caption viewers at a p95 of 36 to 94 ms.+ - [Translation APIs in 2026: Pricing and Features](https://verli.app/blog/choosing-a-translation-api): Translation APIs cost $10 to $27.50 per million characters, and the cheapest LLM we ran about 100 times less. DeepL, Google, Azure, Amazon and LLMs compared.+ - [Zoom voice translator](https://verli.app/zoom-voice-translator)+ - [Teams](https://verli.app/teams)