Addis AI

We benchmarked Addis Voice 2 against six other text-to-speech systems on Amharic. 100 prompts from FLEURS, Waxal and Horn-ASR. Seven systems. Same text across all of them. We measured it two ways: how it sounds to native Amharic listeners, and how accurately a third-party speech recogniser transcribes it. @addisassistantai

Every AI SDK lets you set a system prompt. Addis AI also lets you set a persona: a name, a personality, a brand and speak fluent Amharic and Afan Oromo out of the box. system shapes behaviour. persona shapes identity. One model becomes a bank's support agent, a clinic's triage line, a teacher in your app each with its own voice.

npm install addisai

@addisassistantai

If you're looking to build with Addis AI, here’s where to go depending on what you need. Want to test the voices? 😊Voice Lab → https://addisassistant.com/voice Want to try the models? 😊Playground → https://addisassistant.com/playground Ready to integrate? 😊Docs & APIs → https://docs.addisassistant.com For private cloud, VPC, or on-prem deployments: 😊Enterprise → https://addisai.ch @addisassistantai

Afaan Oromo has 41 million speakers. The frontier models don’t sound native. We gave ElevenLabs, Google, and OpenAI the same script we gave our model.  Sound on. @kalBuilds

We put Addis AI Voice 2 and Gemini 2.5 Flash TTS head-to-head on an Amharic weather report. Which one sounds more natural to you? Try the Tewodros voice on our website. You can start for free. @addisaiassistant

A voice carries more than words. It carries where we come from, the stories we grew up with, the people we love, and the feeling of home. This is what we want our technology to preserve and carry forward. ኢትዮጵያዊነት በፈገግታ ይታያል፣ በድምፅ ይሰማል። ድምፃችን። ድምፃችሁ። @addisassistantai

We are looking for Sidama, Tigrinya, and Wolaita speakers. If you are interested in contributing, join our Discord: 📱https://discord.gg/BqDxzkX3 @b1n1yamBuilds

Payment verification is now back online. Payments should be verified and credits added automatically as usual. Thank you for your patience. If you face any issues, please contact us on Telegram at @Addisai_support @addisassistantai

ቋንቋ እንደ ተፈጥሮ የራሱ የሆነ ሪትምና እስትንፋስ አለው። Language, like nature, has its own rhythm and breath. Made with Addis AI (Voice: Yohannes) 😊addisassistant.com #contentCreation@kalBuilds

Payment verification is currently unavailable due to an issue with the bank’s system. If you’ve made a payment, please DM us your receipt and the email linked to your Addis AI account. We’ll verify it and add the credits manually. We’ll update you once the system is back online. @addisassisstantai

Today we're releasing the Addis AI SDK. Build with speech, voice, language models and translation APIs using a single package. Built for developers building products in African languages from day one. @addisAssistantAI

Research shouldn’t be limited by the language you speak or the keyboard you type on. We’re partnering with @ScholarXIV to bring Addis AI’s speech and real-time models into academic research. Voice dictation, research insights you can listen to, and eventually deeper access to research in Amharic and Afaan Oromo. Excited to show you what we’re building together. @AddisAssistantAI

We're super happy to announce our partnership with Addis AI (@AddisAssistantAI) the leading language model provider for Ethiopian and African languages. Addis AI's text-to-speech, speech-to-text and real-time models are going to help us bring voice dictation, and listening to research insights to our platform - making all the research even more accessible. We can't wait to show you what we're building with this. #Partnerships #AddisAI @ScholarXIV

We gave ElevenLabs, Google, OpenAI, and our model the same Amharic customer support script. Three of them can't greet a customer. 57 million people can hear which three. Sound on. @b1n1yamBuilds

Introducing Addis AI Voices 2. 28 natural, lifelike voices for Amharic and Afaan Oromo, now live through our APIs and platform. addisassistant.com @addisassistantai

Seenaan keenya yeroo afaan ofiitiin himamu @KalBuilds

ሳንተረጉም የምንግባባበት! @addisassistantai

ቋንቋችን ውብ ነው... ድምጻችን ደግሞ ይለያል። ይህ ድምፅ ሙሉ በሙሉ የተፈጠረው በ Addis AI 😊 ጽሑፍን ወደ ድምፅ (Text-to-Speech) በሚቀይረው አዲሱ ስራችን ነው። ማንም ሰው ማይክሮፎን ፊት አልቆመም:: በቅርቡ ሙሉ አገልግሎቱን ይዘን እንመጣለን። አዳዲስ መረጃዎች እንዳያመልጧችሁ ቻናላችንን ይቀላቀሉ! #AddisAIAssistant

Excited to be part of Tech Fest 2026. We’ll be there showcasing what we’ve been building in African language voice AI, including Amharic, Afan Oromo, and a sneak peek at our new Hausa and Swahili ASR models. See you on May 6. @addisassistantai