D.O.T.S AINEWS
Live
OpenAI announces GPT-5 with unprecedented reasoning capabilitiesGoogle DeepMind achieves breakthrough in protein folding for rare diseasesEU passes landmark AI Safety Act with global implicationsAnthropic raises $7B as enterprise demand for Claude surgesMeta open-sources Llama 4 with 1T parameter modelNVIDIA unveils next-gen Blackwell Ultra chips for AI data centersApple integrates on-device AI across entire product lineupSam Altman testifies before Congress on AI regulation frameworkMistral AI reaches $10B valuation after Series C funding roundStability AI launches video generation model rivaling SoraOpenAI announces GPT-5 with unprecedented reasoning capabilitiesGoogle DeepMind achieves breakthrough in protein folding for rare diseasesEU passes landmark AI Safety Act with global implicationsAnthropic raises $7B as enterprise demand for Claude surgesMeta open-sources Llama 4 with 1T parameter modelNVIDIA unveils next-gen Blackwell Ultra chips for AI data centersApple integrates on-device AI across entire product lineupSam Altman testifies before Congress on AI regulation frameworkMistral AI reaches $10B valuation after Series C funding roundStability AI launches video generation model rivaling Sora
Research

Cohere Releases Open-Source Speech Model That Beats OpenAI Whisper and ElevenLabs on Every Benchmark

Cohere has released Transcribe, a 2-billion parameter open-source speech recognition model under the Apache 2.0 license. It posts a 5.42% word error rate — best on the Hugging Face Open ASR Leaderboard — beating OpenAI's Whisper Large v3, ElevenLabs Scribe v2, and Qwen3-ASR-1.7B. It supports 14 languages and is available on Hugging Face and via Cohere's API.

D.O.T.S AI Newsroom

D.O.T.S AI Newsroom

AI News Desk

2 min read
Cohere Releases Open-Source Speech Model That Beats OpenAI Whisper and ElevenLabs on Every Benchmark

Cohere has published Transcribe, an open-source automatic speech recognition model that the company says claims the top position on the Hugging Face Open ASR Leaderboard with a word error rate of 5.42 percent. The benchmark measures how often a model transcribes a word incorrectly — lower is better — and Cohere's figure beats out previous leaders including OpenAI's Whisper Large v3 and ElevenLabs Scribe v2.

Transcribe is a 2 billion parameter model licensed under Apache 2.0, meaning it can be used commercially without restriction. It is available for direct download on Hugging Face and also accessible through Cohere's API and the Model Vault platform. The company is planning to integrate Transcribe into its enterprise AI agent platform, North, in a future update.

The Benchmark Picture

Speech recognition benchmarks measure word error rate (WER) across language diversity, speaker variability, and acoustic conditions. A 5.42% WER at Transcribe's model size — 2 billion parameters — is notable because it achieves competitive accuracy while being smaller than Whisper Large v3. Cohere also claims Transcribe delivers the best throughput among models of similar size, meaning faster transcription at comparable accuracy.

The 14-language support includes English, German, French, and Japanese. While that's narrower than some commercial offerings, it covers the major enterprise use cases and is notably broader than many open-source models at this parameter scale.

Why Cohere Is Doing This

Cohere's go-to-market has historically been enterprise-first, competing with OpenAI and Anthropic in the business API segment rather than the consumer market. Releasing a benchmark-topping open-source model serves two purposes: it establishes Cohere's technical credibility in a domain where OpenAI set the prior standard, and it creates an on-ramp to the Cohere API for developers who start with the open-source version. The Apache 2.0 license removes the licensing friction that has historically slowed enterprise adoption of open models.

CohereSpeech RecognitionOpen SourceWhisperNLPASR
Back to Home

Related Stories

Google's AI Overviews Are Right Nine Times Out of Ten — but the 10% Failure Rate Has a Specific Shape
Research

Google's AI Overviews Are Right Nine Times Out of Ten — but the 10% Failure Rate Has a Specific Shape

A new independent study is the first to systematically measure the factual accuracy of Google's AI Overviews at scale. The headline finding — 90% accuracy — is better than critics expected and worse than Google implies. The more important finding is where that 10% comes from: complex multi-step queries, niche topics, and questions where the web itself is the source of conflicting claims.

D.O.T.S AI Newsroom
Databricks Co-Founder Wins Top Computing Prize — and Says AGI Is 'Already Here'
Research

Databricks Co-Founder Wins Top Computing Prize — and Says AGI Is 'Already Here'

Matei Zaharia, co-founder of Databricks and creator of Apache Spark, has won the ACM Prize in Computing — one of the most prestigious awards in computer science. In interviews accompanying the announcement, Zaharia made a pointed argument: AGI is not a future event but a present condition, and the industry's endless debate about its arrival is obscuring more useful questions about what to do with the AI we already have.

D.O.T.S AI Newsroom
Researchers Fingerprinted 178 AI Models' Writing Styles — and Found Alarming Clone Clusters
Research

Researchers Fingerprinted 178 AI Models' Writing Styles — and Found Alarming Clone Clusters

A new study from Rival analyzed 3,095 standardized responses across 178 AI models, extracting 32-dimension stylometric fingerprints to map which models write like which others. The findings reveal tightly grouped clone clusters across providers — and raise serious questions about whether the AI ecosystem is converging on a single voice.

D.O.T.S AI Newsroom