INDEPENDENT VOICE STUDIO
Use Dessa AI for live voice chat or turn a script into downloadable speech. Choose your voice, check the credit cost, and start from the tools on this page.
Independent service. Not affiliated with Block. Live Talk uses OpenAI GPT-Live-1; speech generation uses OpenRouter.
Looking for Block’s Dessa? Check official accessPress start, allow your microphone, and say hello.
10 credits / minute · rounded up to 6-second blocks
Plan limits: 20 minutes per conversation
Credits are reserved before connecting; unused credits return after server-confirmed settlement. At least 3 credits are required to connect.
Your conversation appears here as you speak.
01 / LISTEN FIRST
Hear what makes Dessa AI interesting before choosing a voice tool. These recordings come from Block’s research release, served from Block’s website. They demonstrate that research model, rather than outputs generated by our Studio. Listen for timing, interruptions, and changes in delivery, then decide which experience you need.
BLOCK RESEARCH RECORDING
Listen for the relationship between laughter, the speaker’s timing, and the next response. This Dessa AI recording is useful for thinking about conversational delivery, rather than testing factual accuracy.
Original recording and synchronized subtitles at BlockBLOCK RESEARCH RECORDING
Follow how the conversation develops when a speaker is searching for a title. Notice when a reply begins and whether it helps the exchange continue. A selected example is not a general knowledge benchmark.
Original recording and synchronized subtitles at BlockBLOCK RESEARCH RECORDING
Compare changes in pacing and vocal expression with the topic of the conversation. Listen through the exchange rather than judging Dessa AI from one emotional reaction or a short isolated excerpt.
Original recording and synchronized subtitles at BlockExplore the Dessa AI sample listening guide for more recordings and listening notes. Public research recordings are selected demonstrations; they do not establish typical performance or guarantee that another voice service will reproduce the same result.
02 / UNDERSTAND THE DIFFERENCE
Dessa AI can refer to Block’s speech research project or to the independent voice tools on dessaai.net. Block describes Dessa as a model trained on speech, able to listen and speak at the same time. That is the research behind the official recordings above. Our website offers a separate browser workflow for live conversation and text to speech, using the providers identified in the Studio.
Use Dessa AI Live Talk to practice speaking or explore an idea through a microphone conversation. Choose a voice, language, and conversation mode before connecting. During a session, you can mute your microphone, read captions, and end the call yourself. Download the transcript when you need a written record.
Turn a prepared script into audio with Dessa AI Text to Speech. Select an available OpenRouter voice, check the character count and credit estimate, then generate a MP3 file. The result stays in the same tool panel for playback and download, so you can review the delivery before using it in your project.
Dessa AI shows the controls for the task you select. Live Talk uses GPT-Live-1; scripted speech uses OpenRouter. Voice availability depends on the provider, and your plan determines session and text limits. Provider configuration and account credits are checked before a call or speech generation can begin.
03 / FROM AN IDEA TO AUDIO
Start with the outcome you want. Dessa AI Live Talk suits an unscripted conversation; Text to Speech suits words you have already written. Both workflows begin in the Studio at the top of this page. Sign up and verify your email to receive 20 free credits. No credit card is required, and both tools share the same trial balance.
Try Dessa AI tutor mode for a short speaking exercise. Set a topic and ask for feedback after your answer. Keep the first conversation focused so you can review the captions afterward.
Use Dessa AI interview mode to rehearse answering aloud. Name the role and request one question at a time. Review your transcript for answers that need a clearer example.
Use Dessa AI Text to Speech for a short introduction, product explanation, or lesson script. Check names and punctuation before generating, then listen to the whole file before publishing.
04 / FIND THE RIGHT EXPERIENCE
Looking for a Dessa AI download, API, or official waitlist? Check the research project’s current access information before subscribing to any independent service. A familiar name does not mean the same model is running behind the interface. Compare the workflow and provider as well as the sound of a selected recording.
| Question | Block’s Dessa | dessaai.net Studio |
|---|---|---|
| What is it? | A speech research model | Browser tools for conversation and scripted speech |
| Where do I start? | Block’s official project and research recordings | Live Talk or Text to Speech on this page |
| Which provider is used? | Block’s own Dessa model | GPT-Live-1 for calls; OpenRouter for TTS |
| Does a Studio plan grant model access? | Research access is managed by Block | A plan supplies credits for this Studio only |
The official-source documentation review dated 2026-10-25 (America/Los_Angeles) records: Public API: No public model API reference was found among the inspected page links. Model weights: No model-weight release artifact or model license was linked from the inspected page. Not verified does not mean unavailable. Check Block’s official source; Studio credits do not grant access to its model.
Block Dessa research record: (America/Los_Angeles). Official Dessa AI research source
Review schedule. Next review due: (America/Los_Angeles). This is an editorial interval, not a guarantee of current availability.
Official Block page text and linked destinations re-read on October 25, 2026 through an AI-assisted documentation review. No human editorial sign-off or interactive model evaluation is claimed. This is a dated documentation record, not a live availability feed. Recheck the official source before relying on access, pricing or compatibility.
Need a consumer app, developer API or local research framework? Explore Dessa AI alternatives by task.
Read the Dessa vs GPT-Live comparison for the listening criteria and product boundaries, or use the Dessa AI access guide to find the official source. Availability can change; follow the dated source rather than assuming a published demo is a released API.
05 / UNDERSTAND THE COST
Dessa AI Studio uses one credit balance for both tools. Live Talk costs one credit per six seconds, or ten credits per minute. Text to Speech costs one credit per 160 trimmed Unicode characters, rounded up for each request. A 161-character script therefore uses two credits. These are our service’s rules, not Block’s research pricing.
660 credits / month
A smaller monthly allowance for occasional conversation practice and short scripts.
20-minute calls · 2,000 characters per script
2,000 credits / month
A larger allowance and longer inputs for repeated practice or narration work.
45-minute calls · 3,500 characters per script
4,100 credits / month
The largest monthly allowance, with the longest supported calls and scripts.
90-minute calls · 5,000 characters per script
Review Dessa AI plans and billing details before purchasing. Generating audio and starting a call require an account, sufficient credits, and an available provider. Listening to public research samples does not require a paid plan. Your remaining balance appears in the Studio before you begin.
06 / BEFORE YOU START
Find the provider, access, and billing details that affect your next step.
No. dessaai.net is an independent voice service. Our Live Talk tool uses GPT-Live-1 and our Text to Speech tool uses OpenRouter. Block develops its own Dessa speech model. Official recordings on this website are identified as research samples and linked to their source; they are not demonstrations of our providers.
Sign up and verify your email to receive 20 free credits once per account, with no credit card required. Use them for up to two minutes of Live Talk or spend them on text-to-speech generation. Both tools share the same credit balance. You can also read the guides and listen to public research recordings for free.
The official-source documentation review dated 2026-10-25 (America/Los_Angeles) records: Public API: No public model API reference was found among the inspected page links. Model weights: No model-weight release artifact or model license was linked from the inspected page. Not verified does not mean unavailable. Check Block’s official source; Studio credits do not grant access to its model.
Completed Text to Speech requests provide a MP3 file for playback and download. Live Talk provides a downloadable text transcript. A conversation transcript is not an audio recording, and the Studio does not offer a download of the underlying voice model. Review generated content before sharing it.
Live Talk is metered at one credit per six seconds. Text to Speech counts trimmed Unicode characters and rounds each request up to the next block of 160. For example, 320 characters use two credits and 321 use three. Monthly allowances and per-request limits depend on your plan.
The Studio displays the error and supports retrying an unresolved operation. Confirmed failed speech generations are refunded. A call’s unused reserved credits are returned when settlement completes. If a request is still unresolved, retry the existing operation rather than assuming it failed and starting another. Contact support with the operation details if recovery does not complete.
Live microphone audio is processed through the configured GPT-Live provider; scripts and selected voices are sent to OpenRouter for speech generation. Account and usage records support authentication and billing. Read our Privacy Policy before submitting personal information, and avoid including information you do not want processed by these services.
Ready to try Dessa AI voice tools?
Choose a live conversation or bring a script. Review your settings and credit cost, then start in the Studio above.
Open the voice Studio