Model Development - The Batch | DeepLearning.AI (Page 2)

Gemini 3.8 Live Extended Thinking achieves 82.6% in performance, shown as the tallest bar on the chart.

Machine Learning Research

Voice, Video, and Reasoning in One: Google’s Gemini 3.8 Live with Extended Thinking

Google released two new speech-to-speech voice models, joining a suddenly crowded field of models built to power voice agents.

Chart ranks Artificial Analysis models by score; Xiaomi's MiMo-V2.6-Pro-RL leads open weights category.

Machine Learning Research

An Unexpected Open Weights Leader: It’s Xiaomi’s MiMo-V2.6-Pro RL and MiMo-V2.6-Flash

Xiaomi, best known for its smartphones and electric vehicles, released the highest-scoring open weights model on Artificial Analysis’ Intelligence Index.

Diagram showing main agent planning and reviewing tasks, while sidekick explores, writes code, and fixes bugs.

Machine Learning Research

One Agent Works, Another Directs: Devin Fusion with SWE-2 rivals Claude Fable 5.1 and GPT-6 Astra output at lower costs

Benchmarking a coding agent typically means scoring one model within one harness.

Jev, marked in pink, stands out in the graph, offering high accuracy for lower cost compared to peers.

Machine Learning Research

Models Built to Do One Thing Well: Jev, a classification model, takes the developer world by storm, inspires imitators

While most companies focus on generative and reasoning models, one company is betting on a class of models that isn’t either.

Bar chart comparing AI models, showing Claude Opus 5.5 with low prompt injection risk, highlighting robustness.

Machine Learning Research

Claude Opus 5.5 Leaps Forward: Anthropic’s updated Opus is the most capable AI model yet, with price reduction on tokens

A week and a half after CEO Dario Amodei proposed slowing down AI development, Anthropic released an AI model that promises to be first in a larger family.

Bar graph showing word error rates of 31 transcription models with Muse Voice at the lowest error rate.

Machine Learning Research

Transcription Battles Heat Up: Google, Meta, and Microsoft all release speech-to-text models suited for agentic input

Speech recognition, a core but sometimes overlooked AI technology in the age of generative models, is having a moment.

Graph trends show steady ECI increase from 120 to 170, highlighting models like Opus and Mythos enhancing capability.

Machine Learning Research

Fable Holds The Top Spot (For Now): Anthropic’s Claude Fable 5.1 is an agentic powerhouse

While last week’s OpenAI launch may have made a bigger splash, Anthropic’s new model remains an agentic workhorse, with top marks on independent evaluations from Artificial Analysis and Vals AI.

A graph shows GPT-6 Astra's accuracy over cost, with stars marking its top performance at lower API costs.

Machine Learning Research

GPT-6 Astra Is a Star: OpenAI’s latest model shines at computer use, front-end engineering, and general knowledge

OpenAI’s new model tops or comes close to topping AI leaderboards, and it does so using a fraction of the tokens and at a fraction of the cost of the few models that outperform it.

A table compares performance metrics of large language models in legal and general domains.

Machine Learning Research

Custom Models for Law, News, and Finance: Thomson Reuters’ Thomson LLM, based on Qwen 3.5, trained on proprietary data

Thomson Reuters launched a proprietary large language model family called Thomson.

The graph situates GLM-5.3-Flash high on performance with low cost, highlighted in red, among many AI models.

Machine Learning Research

Ox Alpha Revealed as GLM-5.3-Flash: Z.ai’s latest model built viral buzz as Ox Alpha while being served on China-made chips

For over a week, the name and maker of the most-used model on OpenRouter remained unknown.

Bar graph compares GPT-5.6 Sol Ultrafast and Sol, highlighting speed differences in processing tasks.

Machine Learning Research

Inside AI’s Need for Speed: OpenAI Partners with Cerebras, Google Releases Gemini 3.7 Flash, Nvidia Debuts Nemotron 3.5 Lightning

Developers who already track models’ cost and accuracy have good new reasons to pay closer attention to a third essential factor: speed.

GLM-5.3 leads in benchmarks with high scores, compared to Kimi K3 and others, showcasing model improvements.

Machine Learning Research

GLM-5.3 Makes Cybersecurity Gains: Z.ai delayed weights for GLM-5.3 due to cybersecurity risk

Z.ai’s latest flagship model effectively ties open-weights leader Kimi K3 on Artificial Analysis’ index of intelligence benchmarks.