Integration with OpenCode and Compatible Tools
QuantenRam APIs are compatible with OpenCode and can be integrated into existing agentic workflows. For most users, OpenCode is the recommended starting point. Those who wish to use a QuantenRam-compatible fork of oh-my-openagent can do so independently.
Recommended Approach
The recommended sequence for new users is:
- Install OpenCode - Start with the standard OpenCode installation
- Configure QuantenRam API - Set up QuantenRam as a provider in OpenCode
- Optional: Use the Fork - Use the QuantenRam-compatible fork of oh-my-openagent if desired
OpenCode Integration
QuantenRam is accessible via standardized OpenAI-compatible APIs. This enables direct integration into OpenCode without proprietary interfaces.
{
"provider": {
"type": "openai-compatible",
"base_url": "https://quantenram.net/v1",
"api_key_env": "QUANTENRAM_API_KEY"
},
"models": {
"default": "quantenram-start/glm-5",
"coding": "quantenram-coding/qwen3codernext",
"reasoning": "quantenram-zenmaster/gpt-5.4"
}
}This configuration shows the typical integration of QuantenRam APIs into OpenCode. Models are addressed via alias IDs that simplify routing and pricing.
QuantenRam-compatible Fork of oh-my-openagent
In addition to direct OpenCode usage, a fork of oh-my-openagent exists that provides QuantenRam-specific configurations and aliases. This fork is maintained independently of QuantenRam's main development and is available under the same license as the original.
Repository
The fork is available at kschlender/oh-my-quantenram on GitHub. It contains preconfigured agent roles and model mappings for QuantenRam APIs.
License
The fork is subject to the same license terms as oh-my-openagent. Usage is free for non-commercial purposes.
Updates
The fork is regularly synchronized with upstream changes but contains no proprietary modifications to the core.
Technical Compatibility
QuantenRam implements the OpenAI API standard for chat completions. This guarantees compatibility with all tools that support this standard.
- API Version: OpenAI API v1 compatible
- Authentication: Bearer Token (QUANTENRAM_API_KEY)
- Streaming: Server-Sent Events (SSE) supported
- Models: Addressable via alias IDs
Separation of Concepts
We deliberately distinguish between three levels in our documentation:
OpenCode
The recommended standard software for agentic workflows. OpenCode is an independent project that works with various providers.
QuantenRam APIs
The API interfaces offered via quantenram.net. They implement the OpenAI standard and are compatible with OpenCode.
QuantenRam-compatible Fork
An optional, additional configuration collection for users who prefer specific QuantenRam presets.
Important: There is no official partnership, approval, or brand authorization between QuantenRam and OpenCode. QuantenRam implements an open standard that is supported by OpenCode and many other tools.
Quick Start with OpenCode
# 1. Install OpenCode
# Follow the official installation instructions at opencode.ai
# 2. Create QuantenRam API Key
# Generate a new API key at https://quantenram.net/keys
# 3. Adjust configuration
# Add the QuantenRam provider in your OpenCode configuration:
{
"provider": {
"type": "openai-compatible",
"base_url": "https://quantenram.net/v1",
"api_key_env": "QUANTENRAM_API_KEY"
}
}
# 4. Set API key
export QUANTENRAM_API_KEY="qr_xxxxxxxxxxxxxxxx"
# 5. Test
opencode chat "Hello World"Frequently Asked Questions
Do I have to use the fork?
No. The fork is optional. OpenCode works directly with QuantenRam APIs once you configure the provider as described above.
Is the fork official?
No. The fork is maintained independently and is available under the same license as the original repository. There is no official affiliation with OpenCode or any other upstream project.
Which models can I use?
All models available via quantenram.net are accessible through OpenCode. They are addressed via alias IDs such as quantenram-start/glm-5 or quantenram-zenmaster/gpt-5.4.
Are there limitations?
QuantenRam implements the OpenAI standard. Features that go beyond this standard (proprietary OpenCode features) may differ or be unavailable.