diff --git a/README.md b/README.md index 7a661b0..eff3cbd 100644 --- a/README.md +++ b/README.md @@ -33,11 +33,8 @@ Google Cloud Platform's DLP API. * text, html, csv, json files * images * PDF files (rendering as images) - * [GCP Vertex AI](https://cloud.google.com/vertex-ai/docs) based redaction - * text, html, csv, json files - * images that are supported by the models - * PDF files (rendering as images) - * [Gemini LLM](https://ai.google.dev/gemini-api/docs) based redaction + * [GCP Vertex AI](https://cloud.google.com/vertex-ai/docs) based redaction using any available models such as + Gemini, Claude, etc.: * text, html, csv, json files * images that are supported by the models * PDF files (rendering as images) @@ -167,6 +164,15 @@ You need to provide the URLs for text analysis and image redaction endpoints usi ### GCP Vertex AI +Vertex AI redacter supports any available models etc on GCP Vertex AI Models Garden, such as: + +- Google Gemini +- Claude +- Google Imagen 3 +- etc. + +Default models are set to Gemini models. + To be able to use GCP Vertex AI you need to: - authenticate using `gcloud auth application-default login` or provide a service account key @@ -184,19 +190,6 @@ In case you have access to native image editing models such as Google Imagen 3, `--gcp-vertex-ai-native-image-support` option. Without native image support, the tool will use LLM output and editing images by coordinates. -### Gemini LLM - -Consider using Vertex AI redacter for more flexibility instead of Gemini LLM. - -To be able to use Gemini as DLP/redacter you need to: - -- authenticate using `gcloud auth application-default login --client-id-file=.json` or provide a - service account key - using `GOOGLE_APPLICATION_CREDENTIALS` environment variable. - Please note that you need to also configure OAuth setup following the - official [instructions](https://ai.google.dev/gemini-api/docs/oauth#set-cloud). -- provide a GCP project id using `--gcp-project-id` option. - ### Open AI LLM To be able to use Open AI LLM you need to provide an API key using `--open-ai-api-key` command line option.