Genspark Tested GPT-Live on 80 Real Calls: Task Completion Doubled, 92% Comprehension
OpenAIDevs · x · 2026-09-11
GPT-Live-1 is now available via API, bringing ChatGPT-style natural voice to apps. Genspark ran it through 80 real restaurant booking calls: task completion more than doubled over the previous generation with 92% perfect comprehension, interruptions are smoother, and it switches to new questions mid-sentence without pause. It's live in Genspark's Call for Me, Realtime Voice Agent, and the new GenTeam, where you can call your agents to hand off work and check results.
More from Models
- Kimi K3.1 Pro tops MathArena open models, closing frontier gap to about 6 months — teortaxesTex · 2026-10-26
- Ruff creator Charlie Marsh: frontier models find infinite bugs in any software — charliermarsh · 2026-10-26
- Dev testing one skill across Opus 5.5 and Sol 6.1 finds instruction following 'all over the map' — lucasmeijer · 2026-10-26
- supermemory unveils learner-1, claiming continual learning for any model — julianweisser · 2026-10-26
- Anthropic says Opus 5.5 cuts cost per task by ~40% vs Opus 5 — taufiqintech · 2026-10-26
- AI search APIs ranked: top 12 within 7 points, yet prices differ nearly 7x — FinanceYF5 · 2026-10-26