Rio de Janeiro has unveiled a competitive AI model, outperforming several leading LLMs, just a day after the U.S
Rio de Janeiro's municipal IT company IplanRIO unveiled Rio 3.5 Open 397B—a 397-billion-parameter model fine-tuned on Alibaba's Qwen—which outperforms DeepSeek V4 and legacy ChatGPT/Claude in benchmarks. Signals that frontier AI competitiveness no longer requires massive training budgets: optimized open-source models now viable for governments and enterprises, reshaping the development landscape.
Rio 3.5 Open 397B has 397 billion parameters
Jun 14, 2026
06:08 pm
What's the storyIn a surprising turn of events, the city government of Rio de Janeiro in Brazil has launched an advanced artificial intelligence (AI) model.
Dubbed Rio 3.5 Open 397B, this 397-billion-parameter AI is already outperforming some of the top large language models (LLMs) in the world.
Rio de Janeiro's IplanRIO released Rio 3.5 Open 397B, a fine-tuned Alibaba Qwen model that outscored DeepSeek V4 Pro, but…
Rio de Janeiro released a frontier-class AI model that claimed to beat Alibaba's best. Then Nex showed up with receipts.
ChatGPT maker OpenAI said its latest powerful artificial intelligence model series will be released to the public on Thursday, as…
GPT-5.6 impresses the critics, but Fidji Simo's exit leaves OpenAI's focus — and its org chart — in flux. PLUS: Meta plays…
Z.ai's GLM-5.2, with a 1 million token context window, surpasses OpenAI's GPT 5.5 on key benchmarks, offering open-source…
This week saw the releases of, among other things:
