Cekura Bench
Speech-to-speech model benchmarks on live phone calls
Get picks like this daily. The day's top launches, AI/tech news, and a weekly opportunity spotlight — straight to your inbox.
This is 1 of 17 launches in algorithm and model benchmarking platforms — see how it stacks up on momentum and crowding →
137 other launches read as similar to this one →
Details
- External ID
- 1271901
- Source
- PH
- Company
- —
- Product
- Cekura
- Website domain
- producthunt.com
- Launched
- Oct. 8, 2026
- Cohort
- —
- Upvotes
- 143
- Upvotes percentile
- 0.99
- Tags
- SaaS, Developer Tools, Artificial Intelligence
Description
Cekura Bench publishes voice AI benchmarks you can verify. Our new speech-to-speech benchmark tests 9 realtime voice models, including GPT Realtime 2.1, Gemini Live, Grok and Phonic, as complete phone agents on live calls: 82 scenarios, three runs each. Models are ranked on reliability, data accuracy, stalled calls, response time and cost, and every call transcript is public. Cekura Bench also covers voice agent benchmarks and STT benchmarks, with TTS benchmarks coming soon.
Enrichment
- Niche
- algorithm and model benchmarking platforms
- Vertical
- Horizontal
- Function
- Observability & eval
- Audience
- Developer
- AI stance
- AI-native
- Project type
- Commercial product
- Normalized one-liner
- benchmarking platform for speech-to-speech models on phone calls
- Manually corrected
- False
Could you build this?
Partial While the reporting dashboard and web UI are straightforward, executing automated, reproducible benchmark evaluations across live telephony networks with low latency and rigorous metrics requires specialized telecom integration and audio analysis infrastructure.
What it would actually take: The stack would involve Twilio or FreeSWITCH/Asterisk SIP trunks running automated calls against target speech-to-speech APIs, recording bidirectional audio streams, and piping them into transcription and analysis pipelines (measuring latency, word error rate, interruption handling). The difficult part is building deterministic multi-turn conversational evaluation scenarios over real PSTN audio with sub-100ms timing precision.
Competitors
Other products that read as similar to this one — 137 launches clear the similarity bar, closest 8 shown.
Attention rank: #5 of 138 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).
Launched 331 days after the earliest competitor.
- Inception Mercury Voice · ph · 2026-10-01 · 1 upvotes · similarity 0.48
- GPT-Live · ph · 2026-07-09 · 211 upvotes · similarity 0.48
- Τ³-Bench is out · hn · 2026-03-25 · 12 upvotes · similarity 0.47
- Realtime TTS-2 · ph · 2026-05-06 · 152 upvotes · similarity 0.46
- Gemini 3.5 Live Translate · ph · 2026-06-10 · 253 upvotes · similarity 0.46
- Live Qwen3-Omni API (open-source speech-to-speech) · hn · 2025-12-02 · 5 upvotes · similarity 0.45
- Audio AI had a wild day · hn · 2026-01-23 · 5 upvotes · similarity 0.44
- SpeechifyAI Simba Voice Agents · ph · 2026-07-13 · 161 upvotes · similarity 0.44