FOSSWire
Search/
Local AI & Edge

Ollama v0.40.5 adds NVIDIA Nemotron 3.5 support and cloud decision models

Brief written by the FOSSWire AI editor from Ollama How we write

1 min read 34 reads

Key Takeaway

Ollama v0.40.5 introduces support for NVIDIA Nemotron 3.5 models and allows running decision models via the ollama.com cloud.

Summary

The latest Ollama release adds support for NVIDIA's Nemotron 3.5 Super and Nemotron 3.5 Bolt vision-language models. It also introduces the ability to run decision models through the ollama.com cloud using specific identifiers, and improves error reporting for cloud access and OpenAI-compatible tool calls.

Technical updates in Ollama v0.40.5

  • Added support for NVIDIA Nemotron 3.5 Super (120B MoE) and Nemotron 3.5 Bolt (74B).
  • Introduced cloud-based decision models via clef:cloud and clef-flash:cloud prefixes.
  • Updated llama.cpp and MLX dependencies to include Nemotron 3.5 support.
  • Fixed an issue where OpenAI-compatible conversations with custom tool calls would fail.
  • Improved error messaging for desktop app sign-in, payment, and ChatGPT gateway access issues.

Why it matters

The inclusion of large-scale MoE models like Nemotron 3.5 expands Ollama's capability for complex reasoning and vision tasks, while cloud integration offers a way to use heavy models without local hardware constraints.

Original reference: github.com →