Gemini¶

Configure HolmesGPT to use Google's Gemini models via Google AI Studio.

Setup¶

Configuration¶

Holmes CLIHolmes Helm ChartRobusta Helm Chart

export GEMINI_API_KEY="your-gemini-api-key"
export TOOL_SCHEMA_NO_PARAM_OBJECT_IF_NO_PARAMS=true
export MODEL="gemini/<your-gemini-model>"
holmes ask "what pods are failing?"

# Or specify the model explicitly per command:
holmes ask "what pods are failing?" --model="gemini/<your-gemini-model>"

Create a Kubernetes secret in the namespace Holmes runs in:

kubectl create secret generic holmes-gemini \
  --from-literal=GEMINI_API_KEY="your-gemini-api-key" \
  -n <namespace>

When using the standalone Holmes Helm Chart, update your values.yaml:

extraEnvVarsSecrets:
  - holmes-gemini

additionalEnvVars:
  - name: TOOL_SCHEMA_NO_PARAM_OBJECT_IF_NO_PARAMS
    value: "true"  # Required for Gemini - see Environment Variables Reference
  # Optional: Set default model (use modelList key name)
  - name: MODEL
    value: "gemini-pro"  # This refers to the key name in modelList below

# Configure at least one model using modelList
modelList:
  gemini-pro:
    api_key: "{{ env.GEMINI_API_KEY }}"
    model: gemini/gemini-pro
    temperature: 1

  gemini-flash:
    api_key: "{{ env.GEMINI_API_KEY }}"
    model: gemini/gemini-1.5-flash
    temperature: 1

  gemini-pro-exp:
    api_key: "{{ env.GEMINI_API_KEY }}"
    model: gemini/gemini-exp-1206
    temperature: 1

Apply the configuration:

helm upgrade holmes robusta/holmes -f values.yaml

Create a Kubernetes secret in the namespace Holmes runs in:

kubectl create secret generic holmes-gemini \
  --from-literal=GEMINI_API_KEY="your-gemini-api-key" \
  -n <namespace>

When using the Robusta Helm Chart (which includes HolmesGPT), update your generated_values.yaml:

holmes:
  extraEnvVarsSecrets:
    - holmes-gemini

  additionalEnvVars:
    - name: TOOL_SCHEMA_NO_PARAM_OBJECT_IF_NO_PARAMS
      value: "true"  # Required for Gemini - see Environment Variables Reference
    # Optional: Set default model (use modelList key name)
    - name: MODEL
      value: "gemini-pro"  # This refers to the key name in modelList below

  # Configure at least one model using modelList
  modelList:
    gemini-pro:
      api_key: "{{ env.GEMINI_API_KEY }}"
      model: gemini/gemini-pro
      temperature: 1

    gemini-flash:
      api_key: "{{ env.GEMINI_API_KEY }}"
      model: gemini/gemini-1.5-flash
      temperature: 1

    gemini-pro-exp:
      api_key: "{{ env.GEMINI_API_KEY }}"
      model: gemini/gemini-exp-1206
      temperature: 1

Apply the configuration:

helm upgrade robusta robusta/robusta -f generated_values.yaml --set clusterName=<YOUR_CLUSTER_NAME>

Using CLI Parameters¶

You can also pass the API key directly as a command-line parameter:

export TOOL_SCHEMA_NO_PARAM_OBJECT_IF_NO_PARAMS=true
holmes ask "what pods are failing?" --model="gemini/<your-gemini-model>" --api-key="your-api-key"

Additional Resources¶

HolmesGPT uses the LiteLLM API to support Gemini provider. Refer to LiteLLM Gemini docs for more details.