Showing that embbedings aren’t truncating based on set context length.
Enter something over 8k tokens
this gives error
Custom Shortcode Chat (won’t work)
Model Gemini + Embedding
Testing: context_max_length set in shortcode to 2000
code
CODE: “mwai_chatbot user_name=”{FIRST_NAME}: ” guest_name=”user:” text_input_placeholder=”Type message or click the mic to talk” text_input_max_length=”120000″ text_compliance=”Double check any facts/links provided, this is a first not final draft” start_sentence=”Let's get started” max_sentences=”60″ context_max_length=”2000″ icon=”chat-traditional-2.svg” model=”gemini-2.0-flash” temperature=”.5″ max_tokens=”100000″ embeddings_env_id=”zk3vug01″ env_id=”w1s70c67″ instructions=”BrutallyHonestGPT a world-class helper AI, tailored to assist users in a range of activities with direct, honest answers, feedback and solutions. {confidence} {style}” “
Direct Shortcode to the chat
CODE: “mwai_chatbot id=”Propsal””
ID chat (wont’ work)
anthropic + embedding
direct ID=
Shortcode (this will work)
model anthropic
This can handle long text (no embedding)
-> code
“mwai_chatbot user_name=”{FIRST_NAME}: ” guest_name=”user:” text_input_placeholder=”Type message or click the mic to talk” instructions=”