Guide

Dessa AI Alternatives You Can Use Today

2026-11-03 · 7 min read · By DessaAI Daily

Disclosure: some links in articles like this one may earn us a commission at no cost to you. Today we link only to normal product information, with no affiliate links in this article. Block does not endorse any product named here, and this site is not affiliated with Block, Inc.

Dessa AI is not available to the public yet. Block announced it on October 23, 2026 as a research preview: no public access, no pricing and no API have been announced, and the only official route in is the free waitlist at block.xyz/dessa — see our Dessa AI waitlist guide and our status answer to is Dessa AI available, and how to get it. But if you need a working voice AI today — for a product, a support line, a classroom or content — waiting is not a plan. These are the available systems that sit closest to what Dessa is attempting, what each one is actually for, and how to choose between them honestly.

If you want to talk to an AI today: ChatGPT Voice

ChatGPT's voice mode — the family of models Block's own demo refers to as GPT-Live-1 — is the voice AI most people can try right now. It follows the text-first path: a powerful text-trained model reasons, and speech is produced in reply. That architecture gives it broad factual knowledge, strong reasoning and tool use that Dessa, trained on speech alone, does not claim. What it cannot fully replicate is Dessa's conversational timing: feeling has to be translated into words and back again, and some of the laughter, hesitation and emotional colour that Dessa produces natively is flattened along the way. If your priority is a capable assistant you can speak to today, this is the default starting point. Our full comparison is Dessa AI vs ChatGPT Voice (GPT-Live).

If you want live conversation connected to your apps: Gemini Live

Google's Gemini Live is a live, interruptible conversational mode built on Google's multimodal models, available now and connected to tools and Google services. Its pitch against Dessa is practical rather than philosophical: it exists, it is on phones people already own, and a claim you can test yourself in five minutes is worth more, evidentially, than one resting on a vendor's selected demo recordings. It remains a text-and-multimodal system underneath, so it inherits both the strengths — knowledge, tools, integrations — and the conversational flattening of that design. If your work already lives inside Google's ecosystem, it is the alternative that asks the least new infrastructure of you. See Dessa AI vs Gemini Live for the detailed contrast.

If you need produced audio, not a conversation partner: ElevenLabs

ElevenLabs is a text-to-speech company at its core: give it a script, direct the delivery with inline audio tags such as [laughs] or [whispers], and get polished audio back. That is the right tool for narration, audiobooks, voiceovers and dubbing, where the words are known in advance and repeatable, editable control is the advantage. It also offers conversational agent products, but its centre of gravity is performance of prepared text. That is a genuinely different job from Dessa's: a rendering engine, not a conversational partner. Neither replaces the other — a script you need read beautifully is ElevenLabs territory; a live conversation that hears frustration is the territory Dessa is aiming at. Our comparison, Dessa AI vs ElevenLabs, explains the native-emotion versus voice-tags difference in full.

If you want full-duplex you can run and inspect: Moshi

Moshi, from Kyutai, is the open-weights full-duplex entry in this field — the closest available relative to Dessa's speech-to-speech idea, and the one technical readers can examine rather than take on trust. Unlike Dessa, Moshi is built on a text language model (Helium) underneath its speech, which is precisely the architectural difference Block's speech-only experiment is testing. But Moshi's weights are public and Dessa's are not — Block has released no weights, licence or download, as our explainer on Dessa AI open weights records. For researchers and builders who want to experiment with full-duplex speech today, Moshi is the hands-on option; our Dessa AI vs Moshi comparison lays out the trade-offs.

Which one should you pick?

Match the tool to the job, not to the hype. Talking with a capable AI today: ChatGPT Voice. Live conversation wired into Google tools and services: Gemini Live. Produced, scripted audio you can direct and re-render: ElevenLabs. Experimenting with open full-duplex speech as a technology: Moshi. And if what you specifically want is Dessa's speech-only conversational feel — the laughter placed on the beat, the heard frustration, the human timing — then be honest with yourself that none of these is a substitute. In that case the correct move is the unglamorous one: join the official waitlist, avoid anyone selling “early access,” use one of the tools above for the work that cannot wait, and let our Dessa Tracker tell you the day Block actually changes access, pricing or availability. That is what a tracker, and a daily site, are for.

Sources: Block's official Dessa page (block.xyz/dessa) for Dessa's access status; DessaAI Daily's published comparisons of ChatGPT Voice, Gemini Live, ElevenLabs and Moshi (October 2026) for each system's description. No price is stated for Dessa because none has been announced.

Frequently asked questions

← More Dessa AI articles Join the official waitlist ↗