Learn Google's Gemini and Anthropic's Claude API with Python

Set up the environment and obtain a Gemini API key for Google Studio, noting regional availability and VPN options. Test the key with curl and review free and paid pricing.

Install the Google generative AI python sdk, set up a virtual environment, and load api keys from a .env file to authenticate to the Gemini API, listing Gemini Pro models.

Explore Gemini, Google's multi-modal AI trained on text, images, audio, video, and code, with Nano, Pro, and Ultra, and see how Gemini Pro via API enables multimodal reasoning and outputs.

Explore Google AI Studio, a browser-based IDE for prototyping with generative models. Create freeform prompts with Gemini Pro Vision, export code in Python, and test text and image inputs.

Explore Google Studio by parameterizing prompts with variables to create dynamic user inputs for Gemini responses. Adjust model parameters like temperature, stop sequences, output length, top k, and top p.

Learn to call the Gemini Pro API with Python to generate text from text inputs, set up authentication, and handle responses with safety filters.

Explore streaming model responses with the Gemini API, delivering text piece by piece to reduce latency, enable chunked output, and improve learner experiences in an LMS.

Learn to generate text from image and text inputs with the Gemini Provision API in Python, including image descriptions and multimodal prompts.

Explore controlling Gemini API responses with a generation config object, adjusting parameters such as candidate count, stop sequence, max output tokens, temperature, top p, and top k for Gemini Pro.

Explore how Gemini and Claude API generation parameters shape outputs by adjusting candidate count, stop sequences, max tokens, temperature, and top k and top p to control creativity and determinism.

Learn to build a multi-turn chat with Gemini using python by applying the start chat method with a history of prompts and responses, sending messages, and inspecting chat history.

Learn to build a custom chatbot with Gemini Pro in Python by starting chat sessions, preserving history, and tailoring the front-end experience and tone for specific applications.

Explore Gemini 1.5 Pro's new features, including audio and video capabilities, a file API, system instructions, and Json mode, and learn authentication, model discovery, and testing with a prompt.

Discover how system instructions in Gemini 1.5 steer model behavior by defining personas, output formats (json or yaml), style, tone, and goals, with practical Python code examples and authentication.

Showcases multimodal prompting with text and image via the Gemini file API. Uses Gemini 1.5 Pro to generate content from a text and image prompt, with 2 GB per file.

Explore how tokens power the Gemini API, including tokenization, vocabulary, and context windows; learn model info, token counts, and multi-modal token handling from images to text.

Learn how to prompt Gemini 1.5 Pro with audio files using the file API to generate transcripts, concise summaries, and emotion-driven analyses with timestamps, unlocking interactive audio prompts.

Discover how Jupyter AI adds generative capabilities to Jupyter Lab and Notebook, boosting productivity, accuracy, and collaboration while simplifying learning with AI coding assistants.

Explore Jupiter AI, an open source project that brings generative artificial intelligence to notebooks via a chat interface and magic commands, and compare GitHub Copilot, Codewhisperer, and Star Coder.

Install and set up Jupyter Lab and Jupyter AI, create a virtual Python environment, manage versions, and ensure OpenAI and Long Chain are ready for Python-based interactive AI experiments.

Learn to run Jupyter AI in Jupyter Lab and connect to providers like OpenAI and Anthropic. Set up API keys and use Jupyter Note to generate code and explanations.

Set up and use Jupyter AI magic commands in notebooks to authenticate with OpenAI, load the Jupyter I magics extension, and list available models across providers.

Explore Jupyter AI in Jupyter Notebook with cell magic, selecting a provider and model by global ID or alias, and using Python basics like sorting lists.

Learn interpolation techniques for advanced prompts, using curly braces and Jupyter references to debug Python code, with examples including list comprehension, triangular numbers, and Hugging Face image models.

Call models from the Hugging Face Hub using Jupiter AI in Jupyter Notebook, authenticate with a read token, and generate a stable diffusion image via a dystopian cyberpunk prompt.

Explore how to build an image-based question-answering app with Gemini Provision API, using Jupyter Notebook and Streamlit to load images, generate code, and create AI-powered recipes or art analyses.

Build a multimodal chat app using the Gemini Provision API in Python, loading the API key via dot env, handling images with Pillow, and iterating prompts in a loop.

Test a Gemini-powered image analysis by uploading an image, asking questions, and receiving a detailed description. Generate a json list of ingredients ranked by allergy risk and explore improvements.

Transform Jupyter notebooks into accessible web apps using Streamlit for a multimodal AI app with the Gemini Provision API, and learn setup with PyCharm and the Gemini SDK for Python.

Build a streamlit web app that uses the Gemini provision API to converse with images. Authenticate with dotenv, upload an image, provide a prompt, and generate a model response.

Save every question and its answer in streamlit session state and display the full chat history in a text area, using a history key to keep a running q&a log.

Develop a photo organization app that uses the Gemini API to analyze images and rename files with meaningful titles, while learning Python techniques like generators and file path handling.

Stream images efficiently from a directory and its subdirectories using a generator that yields one PIL image and its absolute path at a time, for Gemini provisioning.

Rename images in a directory with Gemini Pro to generate descriptive, lowercase, underscore-separated names, while a Python workflow authenticates to Gemini, analyzes each image, and preserves the original extension.

Master prompt engineering to design effective prompts for large language models, guiding tasks with context, using the Gemini API, and leveraging a helper function for high-quality outputs.

Develop effective prompts for LMS by placing clear, specific instructions at the beginning and refining them through feedback and multiple cycles with delimiters to guide the model toward desired output.

Be specific, descriptive, and detailed about the desired outcome, format, and length for prompt engineering, such as an introduction, a bullet list, and a concluding synthesis.

Specify the response format with precise, detailed prompts to control outcome and length. Use an introduction, bullet list, and conclusion to structure the output.

Explore few-shot prompting to guide language models with examples, compare zero-shot and few-shot approaches, and boost prompt results using Gemini and Claude APIs with Python.

Specify the steps required to complete a task to guide the model with a clear sequence, from introducing characters and setting to describing conflict, solution, and resolution.

Explore giving models time to think by crafting longer, specific prompts that provide context, pose tasks to the model, and enable deeper processing for accurate, relevant responses.

Explore tactics for better prompting, including reducing ambiguity, avoiding negations, using positive instructions, and grounding outputs with reliable sources to avoid hallucinations.

Learn to prompt large language models to generate text, summarize articles, and answer questions. Apply few-shot prompts, step-by-step guidance, time to think, positive instructions, and contextual keywords.

Explore the Claude 3 family: Opus, Sonnet, and Haiku, and their multimodal, vision-enabled capabilities with a 200k token context and cost-efficient Haiku.

Set up your environment, install the Anthropic Python SDK, and generate an Anthropic API key to begin making cloud API requests and testing prompts in the workbench.

Learn to generate text from text prompts by making api requests to the cloud with the anthropic python client, including loading the anthropic api key and configuring a model.

Explore how the assistant role guides model responses through teaching by example, and see Gemini generate haiku poems about the sun and moon, illustrating prompt design and response control.

Discover how a system prompt provides context, instructions, and tone to a chat model, and observe how adding one changes responses in practical Python examples.

Learn how streaming Claude responses deliver output in chunks, enabling real-time interaction and reduced latency, with Python code using client.messages.stream for incremental, immediately displayed text.

Learn how multimodal ai vision capabilities let Claude understand and analyze images, describe scenes, and spark creative stories through the chat interface or Python API using standard Claude three models.

explore fetching online images with httpx, handling http status 200, and using the image in a notebook to prompt a model to link visuals to key scientific concepts for kids.

Learn to analyze up to 20 images in a single Claude API request by encoding images to base64, preparing the content for the cloud, and asking visual questions with Python.

Explore how while loops work with the continue statement to skip iterations, print numbers divisible by 13 under 100, and avoid infinite loops caused by missing increments.

Explore the break statement in while loops, exiting the innermost loop to control a lucky-number prompt and the search for prime numbers.

Explore Python list slicing and iteration, from start and stop defaults to stepping and reversing, to concatenation, for loops, and membership tests with in and not in.

Explore Python's list comprehension, part 1, a concise way to build a new list from an iterable, and compare doubling numbers with a for-loop to the one-line approach.

Explore list comprehension with test conditions, filtering numbers divisible by seven, converting numbers to strings, joining them, and finding common elements across lists while handling capitalization differences.

Master dictionaries by using len, adding, removing, and updating key value pairs. Access values with keys and get, handle key errors, and navigate nested lists and dictionaries.

Explore serializing Python objects to json with the built-in json module, using dump and dumps, and understand interoperability and security versus pickle.

Deserialize json into Python objects with json.load and json.loads from files and strings, and understand how json maps to Python types like lists, dictionaries, and tuples.

Practice using Python requests to fetch todos from the json placeholder api, load the json into a Python object, print its type and value. Print only the completed todos.

Install and import requests and json, then fetch data via an http get request, parse json with json.loads, and print only completed tasks from the todos list.

Explore how Streamlit, an open source Python library, lets you create and share web apps for machine learning and data science with fast production deployment.

Discover Streamlit's main concepts: run a Streamlit app, import the library, and build with charts, widgets, and tables, while understanding automatic front-end updates and caching.

Display data with Streamlit using st.write or magic to render text, lists, dictionaries, and dataframes, and explore pandas integration, emoji shortcuts, and static tables via st.dataframe and st.table.

Learn how Streamlit widgets enable interactivity in Python apps, focusing on text input, number input, and a clickable button with labels and return values.

Explore checkbox, radio, and select widgets in Streamlit, including pre-select with index, showing data frames, and using a widget key for session state in interactive apps.

Master the slider, file_uploader, and camera_input widgets to capture, upload, and display files and images. Configure ranges and types, and read CSV, TXT, and Excel with pandas.

Learn to layout a Streamlit app with a left panel sidebar, featuring a country select box and a temperature slider, using layout.py and the command streamlit run layout.py.

Learn how to control app layouts with Streamlit columns, create two or three columns with equal or relative widths, and place widgets side by side, including charts, text, and images.

Learn how to use the expander layout as a multi-element container that users can expand or collapse, and add a bar chart and image using the label syntax.

Learn to display a progress bar in Streamlit during extensive computations by updating a placeholder with each iteration, showing operation in progress, and signaling completion.

Explore Streamlit session state in Python to persist and share widget values across app reruns, manage keys with dictionary-like access, and respond to button and slider interactions.

Explore callbacks in Streamlit using onchange and onclick to synchronize widgets and session state. Build a miles-to-kilometers converter app with columns and interactive inputs.