Global Premiere: Mureka O1 Ushers in the AI Music Revolution

Mureka, a pioneer in AI music generation, proudly announces the groundbreaking O1 Music Generation System, setting a new benchmark for creative expression. With comprehensive multilingual capabilities, scenario-driven background music (BGM) generation, and cutting-edge AI editing features, O1 reshapes the music creation paradigm.

Redefining the Music Creation Paradigm with O1

Multilingual Coverage

• Generate lyrics and music seamlessly across 10 major languages spanning the Americas, Europe, and Asia.
• Covers diverse genres and styles, empowering creators globally.

Scenario-Specific BGM Generation

• Effortlessly craft complete and professionally tailored background music by simply inputting a descriptive scene prompt.

Multi-Track Separation & Download

• Obtain separate tracks for vocals, instrumental accompaniments, and additional layers.
• Enables flexible mixing, detailed editing, and limitless creative opportunities.

Industry-Leading AI Voice Cloning

• State-of-the-art voice cloning technology accurately replicates distinct vocal timbres.
• Quickly generate personalized, high-quality music content with authentic vocal character.

Song Translation (Coming Soon)

• Revolutionary feature: Users upload existing reference songs and input new lyrics in different languages.
• Automatically generates accurate, melody-preserving, cross-language song translation.
• Instantly create personalized versions—transform existing melodies into your own multilingual masterpieces.

Leading Technological Innovations

First-in-Industry Music CoT Architecture

• Incorporates innovative Chain-of-Thought (CoT) technology, marking its pioneering application within music generation.
• Iterative optimization greatly enhances lyrical-melodic coherence, vocal precision, and expressive artistic nuances.

Exceptional Low-Latency Performance

• Through extensive optimization of AI infrastructure, Mureka O1 delivers industry-leading low-latency music generation.
• Offers real-time, premium-quality music generation experiences.

Benchmarking Excellence

Subjective Evaluations:

• Demonstrated superior music quality and overall listener preference compared to industry benchmarks, surpassing competitors in instrumental diversity, arrangement creativity, and vocal richness.

Objective Evaluations:

• Assessed using 100 standardized English prompts to ensure fairness and rigor, leveraging recently published open-source pretrained models, with higher scores indicating better performance:
• Pronunciation Accuracy (WhisperX): accuracy in lyrical pronunciation.
• Music Segment Coherence (All-In-One Music Structure Analyzer): accuracy in music segment reconstruction.
• Text Relevance (CLAP, CLaMP 3): prompt-music alignment.
• Production Quality (Meta Audiobox Aesthetics): user enjoyment, content value, complexity, and overall quality.

Using analysis and testing with the aforementioned popular open-source models, Mureka O1 leads the industry in pronunciation clarity, lyrical accuracy, segment precision, text relevance, and music production quality of generated music.

API Accessibility and Collaborative Opportunities

Advanced Text-to-Music (TTM) API Suite

Mureka offers two versatile APIs tailored specifically for developers and enterprise users:

Standard Music Generation API

• Easily generate diverse styles of music, including instrumental tracks, from simple textual prompts.
• Ideal for creators in digital content, gaming soundtracks, video production, and beyond.

Fine-Tuned Private Library API

• Allows businesses and individuals to upload and fine-tune private music collections (up to 200 tracks).
• Deep model fine-tuning captures nuanced style and melodic preferences, effortlessly creating exclusive branded music and personalized albums—even for users with minimal musical expertise.

Advanced Text-to-Speech (TTS) API Suite

• Benchmark testing reveals that Mureka TTS delivers exceptional performance, surpassing industry leaders such as 11labs, OpenAI, and Microsoft.

• AI Podcast API: Automatically convert scripted dual-person dialogues into complete, natural-sounding podcast episodes.
• Premium Voice Selection API: A curated library of premium, lifelike speaker voices suitable for natural conversational scenarios, customer support, audiobooks, and more.
• Voice Cloning API: Rapidly clone a personalized voice using only a 10-second audio sample—instantly and accurately recreating voice identities.

In conversational scenarios, Mureka TTS achieved an overall listener satisfaction rating of 4.34, consistently ranking among top industry benchmarks.

MusiCoT: A Groundbreaking CoT Methodology for Music Generation

Mureka is excited to unveil MusiCoT (https://MusiCoT.github.io/), an innovative technical breakthrough leveraging Chain-of-Thought (CoT) methodologies. Diverging from conventional autoregressive models that generate audio sequentially, MusiCoT first constructs a comprehensive structural outline of the musical piece prior to predicting individual audio tokens. This revolutionary technique significantly improves structural coherence and precise instrumental arrangements. MusiCoT’s CLAP-model foundation ensures scalability without human annotations, dramatically enhancing interpretability, reliability, and high-fidelity music generation—propelling AI-driven music creation into a structured and highly creative new era.

Experience the revolutionary Mureka O1 system firsthand by visiting Mureka.ai. Join our vibrant Discord community to collaborate with fellow innovators and explore the limitless possibilities in AI-driven music creation.