“`html
- A new option, `thinking 1`, has been added to the `llm-echo` plugin. This allows users to test against LLM versions up to and including 0.32a0.
- The release notes indicate that this feature is intended for automated testing, providing a fake model called “echo” which doesn’t run an actual LLM but instead outputs JSON echoing the prompt as if it were reasoning through it.
- This enhancement enables more comprehensive and flexible testing of applications that use different versions of language models without needing to set up or manage multiple environments or services.
“`
Source Read original →
Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.
