Early reviews of OpenAI dots: best-in-class voice, but feels rushed

OpenAI opened its personal agent product dots for testing from late September to early October, and first feedback from users and developers is sharply polarized: voice capability wins unanimous praise, while the product's maturity is widely criticized as "rushed."

Confirmed

• Voice is the biggest highlight: manosaie's hands-on testing found ChatGPT voice mode still the best voice experience among all competitors; voice combined with tool calling, browser, and computer use is very powerful, opening a new paradigm where AI requires no "active management."
• Feels rushed overall: manosaie explicitly said dots feels "a bit rushed" — it might have wowed at an August launch, but the bar has since been raised; imjustnewatai complained that even websites dots was authorized to log into couldn't be viewed properly.
• Stability issues: SeanOliver had dots log into Google to check trip planning; after taking over the login credentials, the agent went completely unresponsive and then silently kept reading his messages — "read but never replied."
• Interaction design gripes: KPGilmore likes the product itself but dislikes that voice interaction requires "calling dot" like a phone call and waiting to connect, rather than a one-tap instant chat like Codex or ChatGPT Work.
• Strengths and flaws coexist: jdjohnson finds Dot smarter than Grok Bot and Muse, and it self-corrects after mistakes, but it often ignores user-defined skills constraints (e.g., a dedicated workflow for editing a pitch deck).
• Poor multitasking and transparency: davis7 (relayed by RachelVT42), after comparing Dots and Grok Bot, concluded the only credible option in the personal agent space is actually Grok Bot, criticizing Dots for cramming all tasks into a single thread, slow responses, and zero transparency during execution.

Head-to-head comparisons

• Heavy user brandongalang (60+ hours per week in agent harnesses) offered a personal ranking: Muse > Dots > Grok Bot > Instinct, noting it reflects only his own usage patterns.
• Peter Yang concluded after testing that dots and Muse aren't direct competitors: dots and Grok Bot focus on work tasks, while Muse and Instinct lean toward personal life — two different tracks.

Why it matters

dots marks OpenAI's formal entry into the personal agent space. First reviews confirm its voice + tool-calling combo is genuinely ahead, but engineering completeness (login authorization, stability, multitasking, transparency) clearly lags. With competitors like Muse, Grok Bot, and Instinct moving fast, whether dots can close the gaps within its launch window will directly shape the personal agent market landscape.

2026-09-30 ~ 2026-10-02 · 12 related posts

• Hands-on with Dot: smarter than Grok Bot but keeps ignoring custom skill constraints — jdjohnson · 2026-09-30
• [source] OpenAI dots went silent after taking user's Google login, tester reports — SeanOliver · 2026-09-30
• [source] Power User Ranks Personal Agents — Muse Over Dots and Grok Bot — After 60+ Hours a Week in Agent Harnesses — brandon_galang · 2026-10-01
• Early user: OpenAI's rushed Dot can't even view the websites it signs into — imjustnewatai · 2026-10-01
• Peter Yang on OpenAI dots vs. Grok Bot and Muse: two separate races for personal agents — petergyang · 2026-10-01
• [source] OpenAI dots first impressions: best-in-class voice, but the product feels rushed — manosaie · 2026-10-02
• Early impressions: OpenAI dots feels rushed and underwhelming, but voice mode still leads — manosaie · 2026-10-02
• Early hands-on: OpenAI dot feels rushed but voice + tool use impresses — manosaie · 2026-10-02
• Agent tester: dots targets work, Muse targets personal life, they don't compete — AccBalanced · 2026-10-02
• Early OpenAI 'dot' user gripes: voice mode requiring a 'call' feels odd — KPGilmore · 2026-10-02
• Early users pan OpenAI's Dots agent: single thread, very slow, zero visibility vs Grok Bot — RachelVT42 · 2026-10-02

1 near-duplicate retellings: SeanOliver