Ollama Coding Models and Integrations Update

Ollama has released cloud access to the GLM-4.6 and Qwen3-Coder-480B models, while updating the Qwen3-Coder-30B model for improved tool calling. These updates expand the availability of high-performance coding models through Ollama's cloud service and local deployment options.

New Coding Models and Availability

Ollama now provides access to several high-capacity coding models with varying deployment options:

  • GLM-4.6: Available via Ollama's cloud service using the command ollama run glm-4.6:cloud.
  • Qwen3-Coder-480B: Available via the cloud service using ollama run qwen3-coder:480b-cloud. For users with more than 300GB of VRAM, this model is also available for local deployment as qwen3-coder:480b.
  • Qwen3-Coder-30B: This model is available locally via ollama run qwen3-coder:30b and has been updated in Ollama's new engine to provide faster and more reliable tool calling.

IDE and Tool Integrations

Ollama has expanded its support for popular coding tools, allowing users to pull cloud models and integrate them directly into their development environments:

VS Code

To use GLM-4.6 or Qwen3-Coder-480B in VS Code, users must first pull the cloud models (ollama pull glm-4.6:cloud and ollama pull qwen3-coder:480b-cloud). Integration is managed through the Copilot chat sidebar by selecting Ollama as the provider in the Manage models menu.

Zed

Zed (now available for Windows) supports Ollama integration. After pulling the cloud models, users can configure the LLM provider to Ollama and connect using the host URL http://localhost:11434 via the agent panel.

Droid

Droid integration requires installing the Droid CLI and adding a specific configuration to the ~/.factory/config.json file. This configuration defines the model_display_name, base_url (http://localhost:11434/v1), and max_tokens (16,384) for both GLM-4.6 and Qwen3-Coder-480B.

Cloud API Access

Cloud models, including glm-4.6 and qwen3-coder:480b, are accessible directly via the ollama.com cloud API. This requires creating an API key in the user settings and setting the OLLAMA_API_KEY environment variable. Requests are made to the https://ollama.com/api/chat endpoint using a Bearer token for authorization.

Supported Integration Ecosystem

Ollama's official documentation now includes dedicated integration guides for the following tools:

  • Codex
  • Cline
  • VS Code
  • Zed
  • Droid
  • Roo code

Sources

Related

  • Dispatch
    Ollama launch command release

    Ollama has introduced the 'ollama launch' command in version 0.15+, allowing users to set up and run coding tools like Claude Code and OpenCode without manual environment variables or configuration files.

  • Dispatch
    Ollama Cloud Models Preview

    Ollama has introduced Cloud models in preview, enabling users to run large-scale models on datacenter-grade hardware while maintaining the same local toolset and API integration.

  • Dispatch
    Ollama 0.19 Preview: MLX Integration for Apple Silicon Acceleration

    Ollama 0.19 introduces a preview integration with Apple's MLX framework, significantly accelerating performance on Apple Silicon devices through unified memory optimization and NVFP4 support.

  • Dispatch
    Ollama Multimodal Engine Release

    Ollama has introduced a new multimodal engine to improve the reliability and accuracy of local inference for vision models, supporting Llama 4, Gemma 3, Qwen 2.5 VL, and Mistral Small 3.1.

  • Dispatch
    Ollama 0.30 Release Adds GGUF Model Support and Up to 20% Faster NVIDIA Performance

    Ollama 0.30 introduces GGUF model compatibility via llama.cpp and delivers up to 20% faster throughput on NVIDIA GPUs, expanding hardware support through default Vulkan acceleration.