diff --git a/docs/source/users/index.md b/docs/source/users/index.md
index 128ef3b23..50420f716 100644
--- a/docs/source/users/index.md
+++ b/docs/source/users/index.md
@@ -765,6 +765,28 @@ The `--response-path` option is a [JSONPath](https://goessner.net/articles/JsonP
 You can specify an allowlist, to only allow only a certain list of providers, or
 a blocklist, to block some providers.
 
+### Configuring default models and API keys
+
+This configuration allows for setting a default language and embedding models, and their corresponding API keys.
+These values are offered as a starting point for users, so they don't have to select the models and API keys, however,
+the selections they make in the settings panel will take precedence over these values.
+
+Specify default language model
+```bash
+jupyter lab --AiExtension.default_language_model=bedrock-chat:anthropic.claude-v2
+```
+
+Specify default embedding model
+```bash
+jupyter lab --AiExtension.default_embeddings_model=bedrock:amazon.titan-embed-text-v1
+```
+
+Specify default API keys
+```bash
+jupyter lab --AiExtension.default_api_keys={'OPENAI_API_KEY': 'sk-abcd'}
+```
+
+
 ### Blocklisting providers
 
 This configuration allows for blocking specific providers in the settings panel.
diff --git a/examples/commands.ipynb b/examples/commands.ipynb
index b640a7001..0b0c6816c 100644
--- a/examples/commands.ipynb
+++ b/examples/commands.ipynb
@@ -146,8 +146,8 @@
        "| `gpt4all` | Not applicable. | <abbr title=\"Not applicable\">N/A</abbr> | `gpt4all:ggml-gpt4all-j-v1.2-jazzy`, `gpt4all:ggml-gpt4all-j-v1.3-groovy`, `gpt4all:ggml-gpt4all-l13b-snoozy`, `gpt4all:mistral-7b-openorca.Q4_0`, `gpt4all:mistral-7b-instruct-v0.1.Q4_0`, `gpt4all:gpt4all-falcon-q4_0`, `gpt4all:wizardlm-13b-v1.2.Q4_0`, `gpt4all:nous-hermes-llama2-13b.Q4_0`, `gpt4all:gpt4all-13b-snoozy-q4_0`, `gpt4all:mpt-7b-chat-merges-q4_0`, `gpt4all:orca-mini-3b-gguf2-q4_0`, `gpt4all:starcoder-q4_0`, `gpt4all:rift-coder-v0-7b-q4_0`, `gpt4all:em_german_mistral_v01.Q4_0` |\n",
        "| `huggingface_hub` | `HUGGINGFACEHUB_API_TOKEN` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | See [https://huggingface.co/models](https://huggingface.co/models) for a list of models. Pass a model's repository ID as the model ID; for example, `huggingface_hub:ExampleOwner/example-model`. |\n",
        "| `openai` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai:babbage-002`, `openai:davinci-002`, `openai:gpt-3.5-turbo-instruct` |\n",
-       "| `openai-chat` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai-chat:gpt-3.5-turbo`, `openai-chat:gpt-3.5-turbo-1106`, `openai-chat:gpt-3.5-turbo-16k`, `openai-chat:gpt-3.5-turbo-0301`, `openai-chat:gpt-3.5-turbo-0613`, `openai-chat:gpt-3.5-turbo-16k-0613`, `openai-chat:gpt-4`, `openai-chat:gpt-4-0613`, `openai-chat:gpt-4-32k`, `openai-chat:gpt-4-32k-0613`, `openai-chat:gpt-4-1106-preview` |\n",
-       "| `qianfan` | `QIANFAN_AK`, `QIANFAN_SK` | <abbr title=\"You have set all of these environment variables, so you can use this provider's models.\">✅</abbr> | `qianfan:ERNIE-Bot`, `qianfan:ERNIE-Bot-4` |\n",
+       "| `openai-chat` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai-chat:gpt-3.5-turbo`, `openai-chat:gpt-3.5-turbo-0301`, `openai-chat:gpt-3.5-turbo-0613`, `openai-chat:gpt-3.5-turbo-1106`, `openai-chat:gpt-3.5-turbo-16k`, `openai-chat:gpt-3.5-turbo-16k-0613`, `openai-chat:gpt-4`, `openai-chat:gpt-4-0613`, `openai-chat:gpt-4-32k`, `openai-chat:gpt-4-32k-0613`, `openai-chat:gpt-4-1106-preview` |\n",
+       "| `qianfan` | `QIANFAN_AK`, `QIANFAN_SK` | <abbr title=\"You have not set all of these environment variables, so you cannot use this provider's models.\">❌</abbr> | `qianfan:ERNIE-Bot`, `qianfan:ERNIE-Bot-4` |\n",
        "| `sagemaker-endpoint` | Not applicable. | <abbr title=\"Not applicable\">N/A</abbr> | Specify an endpoint name as the model ID. In addition, you must specify a region name, request schema, and response path. For more information, see the documentation about [SageMaker endpoints deployment](https://docs.aws.amazon.com/sagemaker/latest/dg/realtime-endpoints-deployment.html) and about [using magic commands with SageMaker endpoints](https://jupyter-ai.readthedocs.io/en/latest/users/index.html#using-magic-commands-with-sagemaker-endpoints). |\n",
        "\n",
        "Aliases and custom commands:\n",
@@ -252,10 +252,10 @@
        "openai-chat\n",
        "Requires environment variable: OPENAI_API_KEY (set)\n",
        "* openai-chat:gpt-3.5-turbo\n",
-       "* openai-chat:gpt-3.5-turbo-1106\n",
-       "* openai-chat:gpt-3.5-turbo-16k\n",
        "* openai-chat:gpt-3.5-turbo-0301\n",
        "* openai-chat:gpt-3.5-turbo-0613\n",
+       "* openai-chat:gpt-3.5-turbo-1106\n",
+       "* openai-chat:gpt-3.5-turbo-16k\n",
        "* openai-chat:gpt-3.5-turbo-16k-0613\n",
        "* openai-chat:gpt-4\n",
        "* openai-chat:gpt-4-0613\n",
@@ -264,7 +264,7 @@
        "* openai-chat:gpt-4-1106-preview\n",
        "\n",
        "qianfan\n",
-       "Requires environment variables: QIANFAN_AK (set), QIANFAN_SK (set)\n",
+       "Requires environment variables: QIANFAN_AK (not set), QIANFAN_SK (not set)\n",
        "* qianfan:ERNIE-Bot\n",
        "* qianfan:ERNIE-Bot-4\n",
        "\n",
@@ -379,8 +379,8 @@
        "| `gpt4all` | Not applicable. | <abbr title=\"Not applicable\">N/A</abbr> | `gpt4all:ggml-gpt4all-j-v1.2-jazzy`, `gpt4all:ggml-gpt4all-j-v1.3-groovy`, `gpt4all:ggml-gpt4all-l13b-snoozy`, `gpt4all:mistral-7b-openorca.Q4_0`, `gpt4all:mistral-7b-instruct-v0.1.Q4_0`, `gpt4all:gpt4all-falcon-q4_0`, `gpt4all:wizardlm-13b-v1.2.Q4_0`, `gpt4all:nous-hermes-llama2-13b.Q4_0`, `gpt4all:gpt4all-13b-snoozy-q4_0`, `gpt4all:mpt-7b-chat-merges-q4_0`, `gpt4all:orca-mini-3b-gguf2-q4_0`, `gpt4all:starcoder-q4_0`, `gpt4all:rift-coder-v0-7b-q4_0`, `gpt4all:em_german_mistral_v01.Q4_0` |\n",
        "| `huggingface_hub` | `HUGGINGFACEHUB_API_TOKEN` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | See [https://huggingface.co/models](https://huggingface.co/models) for a list of models. Pass a model's repository ID as the model ID; for example, `huggingface_hub:ExampleOwner/example-model`. |\n",
        "| `openai` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai:babbage-002`, `openai:davinci-002`, `openai:gpt-3.5-turbo-instruct` |\n",
-       "| `openai-chat` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai-chat:gpt-3.5-turbo`, `openai-chat:gpt-3.5-turbo-1106`, `openai-chat:gpt-3.5-turbo-16k`, `openai-chat:gpt-3.5-turbo-0301`, `openai-chat:gpt-3.5-turbo-0613`, `openai-chat:gpt-3.5-turbo-16k-0613`, `openai-chat:gpt-4`, `openai-chat:gpt-4-0613`, `openai-chat:gpt-4-32k`, `openai-chat:gpt-4-32k-0613`, `openai-chat:gpt-4-1106-preview` |\n",
-       "| `qianfan` | `QIANFAN_AK`, `QIANFAN_SK` | <abbr title=\"You have set all of these environment variables, so you can use this provider's models.\">✅</abbr> | `qianfan:ERNIE-Bot`, `qianfan:ERNIE-Bot-4` |\n",
+       "| `openai-chat` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai-chat:gpt-3.5-turbo`, `openai-chat:gpt-3.5-turbo-0301`, `openai-chat:gpt-3.5-turbo-0613`, `openai-chat:gpt-3.5-turbo-1106`, `openai-chat:gpt-3.5-turbo-16k`, `openai-chat:gpt-3.5-turbo-16k-0613`, `openai-chat:gpt-4`, `openai-chat:gpt-4-0613`, `openai-chat:gpt-4-32k`, `openai-chat:gpt-4-32k-0613`, `openai-chat:gpt-4-1106-preview` |\n",
+       "| `qianfan` | `QIANFAN_AK`, `QIANFAN_SK` | <abbr title=\"You have not set all of these environment variables, so you cannot use this provider's models.\">❌</abbr> | `qianfan:ERNIE-Bot`, `qianfan:ERNIE-Bot-4` |\n",
        "| `sagemaker-endpoint` | Not applicable. | <abbr title=\"Not applicable\">N/A</abbr> | Specify an endpoint name as the model ID. In addition, you must specify a region name, request schema, and response path. For more information, see the documentation about [SageMaker endpoints deployment](https://docs.aws.amazon.com/sagemaker/latest/dg/realtime-endpoints-deployment.html) and about [using magic commands with SageMaker endpoints](https://jupyter-ai.readthedocs.io/en/latest/users/index.html#using-magic-commands-with-sagemaker-endpoints). |\n",
        "\n",
        "Aliases and custom commands:\n",
@@ -486,10 +486,10 @@
        "openai-chat\n",
        "Requires environment variable: OPENAI_API_KEY (set)\n",
        "* openai-chat:gpt-3.5-turbo\n",
-       "* openai-chat:gpt-3.5-turbo-1106\n",
-       "* openai-chat:gpt-3.5-turbo-16k\n",
        "* openai-chat:gpt-3.5-turbo-0301\n",
        "* openai-chat:gpt-3.5-turbo-0613\n",
+       "* openai-chat:gpt-3.5-turbo-1106\n",
+       "* openai-chat:gpt-3.5-turbo-16k\n",
        "* openai-chat:gpt-3.5-turbo-16k-0613\n",
        "* openai-chat:gpt-4\n",
        "* openai-chat:gpt-4-0613\n",
@@ -498,7 +498,7 @@
        "* openai-chat:gpt-4-1106-preview\n",
        "\n",
        "qianfan\n",
-       "Requires environment variables: QIANFAN_AK (set), QIANFAN_SK (set)\n",
+       "Requires environment variables: QIANFAN_AK (not set), QIANFAN_SK (not set)\n",
        "* qianfan:ERNIE-Bot\n",
        "* qianfan:ERNIE-Bot-4\n",
        "\n",
@@ -537,7 +537,7 @@
     {
      "data": {
       "text/markdown": [
-       "As an AI developed by OpenAI, I'm sorry but I can't provide the information you're asking for because your instruction lacks specific details. Could you please provide more context or details?"
+       "Sorry, I can't provide the information you're asking for because your question lacks specific details. Could you please provide more context or information?"
       ],
       "text/plain": [
        "<IPython.core.display.Markdown object>"
@@ -595,27 +595,15 @@
     {
      "data": {
       "text/markdown": [
-       " This means no HTML, tables, images or other formatting. If you generate output, you must use Markdown. See the Markdown Syntax for more information.\n",
+       " I need someone to enter data from a pdf into excel.\n",
        "\n",
-       "What do you mean by a programming language?\n",
+       "We are looking for an experienced freelancer with very high attention to detail to assist us with a number of tasks. Work includes entering data from a pdf into excel, setting up email template, uploading documents, and general administrative support, such as updating excel sheets with current prices. This is a long-term position. Please provide samples of your work. Please note that we will only accept ...\n",
        "\n",
-       "A programming language is a formal, scripted language used in a computer system to program subroutines for the system. Programming languages are often used because it is more convenient to program a computer in the programming language than in the programming language itself.\n",
+       "I have a PDF form which I want to extract the text from the forms fields and place it in a word file. The form is in French and the extracted text must be translated to English.\n",
        "\n",
-       "What is the difference between a programming language and a scripting language?\n",
+       "I have a PDF file which I want to extract the text from the forms fields and place it in a word file. The form is in French and the extracted text must be translated to English.\n",
        "\n",
-       "A scripting language is a programming language designed to enable a user to create applications that are specific to a particular application, such as a web browser or word processing application. … Languages designed for general use are called scripting languages, and they are also known as a scripting languages.\n",
-       "\n",
-       "Can you use Python to program?\n",
-       "\n",
-       "Python is a high-level programming language. … Many developers and programmers use Python to create applications and make use of its functionality. By building applications in Python, you can also become more familiar with the language.\n",
-       "\n",
-       "What are the 2 types of programming languages?\n",
-       "\n",
-       "A programming language is a set of rules that can be used to write a computer program. The two most common classification systems for computer languages are procedural and object-oriented.\n",
-       "\n",
-       "What is the difference between Python and C?\n",
-       "\n",
-       "C is"
+       "I have a PDF form which I want to extract the text from the forms fields and place it in a word file. The form is in French and the extracted text must be translated to English."
       ],
       "text/plain": [
        "<IPython.core.display.Markdown object>"
@@ -661,8 +649,8 @@
        "| `gpt4all` | Not applicable. | <abbr title=\"Not applicable\">N/A</abbr> | `gpt4all:ggml-gpt4all-j-v1.2-jazzy`, `gpt4all:ggml-gpt4all-j-v1.3-groovy`, `gpt4all:ggml-gpt4all-l13b-snoozy`, `gpt4all:mistral-7b-openorca.Q4_0`, `gpt4all:mistral-7b-instruct-v0.1.Q4_0`, `gpt4all:gpt4all-falcon-q4_0`, `gpt4all:wizardlm-13b-v1.2.Q4_0`, `gpt4all:nous-hermes-llama2-13b.Q4_0`, `gpt4all:gpt4all-13b-snoozy-q4_0`, `gpt4all:mpt-7b-chat-merges-q4_0`, `gpt4all:orca-mini-3b-gguf2-q4_0`, `gpt4all:starcoder-q4_0`, `gpt4all:rift-coder-v0-7b-q4_0`, `gpt4all:em_german_mistral_v01.Q4_0` |\n",
        "| `huggingface_hub` | `HUGGINGFACEHUB_API_TOKEN` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | See [https://huggingface.co/models](https://huggingface.co/models) for a list of models. Pass a model's repository ID as the model ID; for example, `huggingface_hub:ExampleOwner/example-model`. |\n",
        "| `openai` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai:babbage-002`, `openai:davinci-002`, `openai:gpt-3.5-turbo-instruct` |\n",
-       "| `openai-chat` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai-chat:gpt-3.5-turbo`, `openai-chat:gpt-3.5-turbo-1106`, `openai-chat:gpt-3.5-turbo-16k`, `openai-chat:gpt-3.5-turbo-0301`, `openai-chat:gpt-3.5-turbo-0613`, `openai-chat:gpt-3.5-turbo-16k-0613`, `openai-chat:gpt-4`, `openai-chat:gpt-4-0613`, `openai-chat:gpt-4-32k`, `openai-chat:gpt-4-32k-0613`, `openai-chat:gpt-4-1106-preview` |\n",
-       "| `qianfan` | `QIANFAN_AK`, `QIANFAN_SK` | <abbr title=\"You have set all of these environment variables, so you can use this provider's models.\">✅</abbr> | `qianfan:ERNIE-Bot`, `qianfan:ERNIE-Bot-4` |\n",
+       "| `openai-chat` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai-chat:gpt-3.5-turbo`, `openai-chat:gpt-3.5-turbo-0301`, `openai-chat:gpt-3.5-turbo-0613`, `openai-chat:gpt-3.5-turbo-1106`, `openai-chat:gpt-3.5-turbo-16k`, `openai-chat:gpt-3.5-turbo-16k-0613`, `openai-chat:gpt-4`, `openai-chat:gpt-4-0613`, `openai-chat:gpt-4-32k`, `openai-chat:gpt-4-32k-0613`, `openai-chat:gpt-4-1106-preview` |\n",
+       "| `qianfan` | `QIANFAN_AK`, `QIANFAN_SK` | <abbr title=\"You have not set all of these environment variables, so you cannot use this provider's models.\">❌</abbr> | `qianfan:ERNIE-Bot`, `qianfan:ERNIE-Bot-4` |\n",
        "| `sagemaker-endpoint` | Not applicable. | <abbr title=\"Not applicable\">N/A</abbr> | Specify an endpoint name as the model ID. In addition, you must specify a region name, request schema, and response path. For more information, see the documentation about [SageMaker endpoints deployment](https://docs.aws.amazon.com/sagemaker/latest/dg/realtime-endpoints-deployment.html) and about [using magic commands with SageMaker endpoints](https://jupyter-ai.readthedocs.io/en/latest/users/index.html#using-magic-commands-with-sagemaker-endpoints). |\n",
        "\n",
        "Aliases and custom commands:\n",
@@ -768,10 +756,10 @@
        "openai-chat\n",
        "Requires environment variable: OPENAI_API_KEY (set)\n",
        "* openai-chat:gpt-3.5-turbo\n",
-       "* openai-chat:gpt-3.5-turbo-1106\n",
-       "* openai-chat:gpt-3.5-turbo-16k\n",
        "* openai-chat:gpt-3.5-turbo-0301\n",
        "* openai-chat:gpt-3.5-turbo-0613\n",
+       "* openai-chat:gpt-3.5-turbo-1106\n",
+       "* openai-chat:gpt-3.5-turbo-16k\n",
        "* openai-chat:gpt-3.5-turbo-16k-0613\n",
        "* openai-chat:gpt-4\n",
        "* openai-chat:gpt-4-0613\n",
@@ -780,7 +768,7 @@
        "* openai-chat:gpt-4-1106-preview\n",
        "\n",
        "qianfan\n",
-       "Requires environment variables: QIANFAN_AK (set), QIANFAN_SK (set)\n",
+       "Requires environment variables: QIANFAN_AK (not set), QIANFAN_SK (not set)\n",
        "* qianfan:ERNIE-Bot\n",
        "* qianfan:ERNIE-Bot-4\n",
        "\n",
@@ -857,8 +845,8 @@
        "| `gpt4all` | Not applicable. | <abbr title=\"Not applicable\">N/A</abbr> | `gpt4all:ggml-gpt4all-j-v1.2-jazzy`, `gpt4all:ggml-gpt4all-j-v1.3-groovy`, `gpt4all:ggml-gpt4all-l13b-snoozy`, `gpt4all:mistral-7b-openorca.Q4_0`, `gpt4all:mistral-7b-instruct-v0.1.Q4_0`, `gpt4all:gpt4all-falcon-q4_0`, `gpt4all:wizardlm-13b-v1.2.Q4_0`, `gpt4all:nous-hermes-llama2-13b.Q4_0`, `gpt4all:gpt4all-13b-snoozy-q4_0`, `gpt4all:mpt-7b-chat-merges-q4_0`, `gpt4all:orca-mini-3b-gguf2-q4_0`, `gpt4all:starcoder-q4_0`, `gpt4all:rift-coder-v0-7b-q4_0`, `gpt4all:em_german_mistral_v01.Q4_0` |\n",
        "| `huggingface_hub` | `HUGGINGFACEHUB_API_TOKEN` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | See [https://huggingface.co/models](https://huggingface.co/models) for a list of models. Pass a model's repository ID as the model ID; for example, `huggingface_hub:ExampleOwner/example-model`. |\n",
        "| `openai` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai:babbage-002`, `openai:davinci-002`, `openai:gpt-3.5-turbo-instruct` |\n",
-       "| `openai-chat` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai-chat:gpt-3.5-turbo`, `openai-chat:gpt-3.5-turbo-1106`, `openai-chat:gpt-3.5-turbo-16k`, `openai-chat:gpt-3.5-turbo-0301`, `openai-chat:gpt-3.5-turbo-0613`, `openai-chat:gpt-3.5-turbo-16k-0613`, `openai-chat:gpt-4`, `openai-chat:gpt-4-0613`, `openai-chat:gpt-4-32k`, `openai-chat:gpt-4-32k-0613`, `openai-chat:gpt-4-1106-preview` |\n",
-       "| `qianfan` | `QIANFAN_AK`, `QIANFAN_SK` | <abbr title=\"You have set all of these environment variables, so you can use this provider's models.\">✅</abbr> | `qianfan:ERNIE-Bot`, `qianfan:ERNIE-Bot-4` |\n",
+       "| `openai-chat` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai-chat:gpt-3.5-turbo`, `openai-chat:gpt-3.5-turbo-0301`, `openai-chat:gpt-3.5-turbo-0613`, `openai-chat:gpt-3.5-turbo-1106`, `openai-chat:gpt-3.5-turbo-16k`, `openai-chat:gpt-3.5-turbo-16k-0613`, `openai-chat:gpt-4`, `openai-chat:gpt-4-0613`, `openai-chat:gpt-4-32k`, `openai-chat:gpt-4-32k-0613`, `openai-chat:gpt-4-1106-preview` |\n",
+       "| `qianfan` | `QIANFAN_AK`, `QIANFAN_SK` | <abbr title=\"You have not set all of these environment variables, so you cannot use this provider's models.\">❌</abbr> | `qianfan:ERNIE-Bot`, `qianfan:ERNIE-Bot-4` |\n",
        "| `sagemaker-endpoint` | Not applicable. | <abbr title=\"Not applicable\">N/A</abbr> | Specify an endpoint name as the model ID. In addition, you must specify a region name, request schema, and response path. For more information, see the documentation about [SageMaker endpoints deployment](https://docs.aws.amazon.com/sagemaker/latest/dg/realtime-endpoints-deployment.html) and about [using magic commands with SageMaker endpoints](https://jupyter-ai.readthedocs.io/en/latest/users/index.html#using-magic-commands-with-sagemaker-endpoints). |\n",
        "\n",
        "Aliases and custom commands:\n",
@@ -963,10 +951,10 @@
        "openai-chat\n",
        "Requires environment variable: OPENAI_API_KEY (set)\n",
        "* openai-chat:gpt-3.5-turbo\n",
-       "* openai-chat:gpt-3.5-turbo-1106\n",
-       "* openai-chat:gpt-3.5-turbo-16k\n",
        "* openai-chat:gpt-3.5-turbo-0301\n",
        "* openai-chat:gpt-3.5-turbo-0613\n",
+       "* openai-chat:gpt-3.5-turbo-1106\n",
+       "* openai-chat:gpt-3.5-turbo-16k\n",
        "* openai-chat:gpt-3.5-turbo-16k-0613\n",
        "* openai-chat:gpt-4\n",
        "* openai-chat:gpt-4-0613\n",
@@ -975,7 +963,7 @@
        "* openai-chat:gpt-4-1106-preview\n",
        "\n",
        "qianfan\n",
-       "Requires environment variables: QIANFAN_AK (set), QIANFAN_SK (set)\n",
+       "Requires environment variables: QIANFAN_AK (not set), QIANFAN_SK (not set)\n",
        "* qianfan:ERNIE-Bot\n",
        "* qianfan:ERNIE-Bot-4\n",
        "\n",
@@ -1045,7 +1033,7 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "{'product': 'colorful socks', 'text': ' Chroma Sox'}\n"
+      "{'product': 'colorful socks', 'text': ' Chroma Socks '}\n"
      ]
     }
    ],
@@ -1103,8 +1091,8 @@
        "| `gpt4all` | Not applicable. | <abbr title=\"Not applicable\">N/A</abbr> | `gpt4all:ggml-gpt4all-j-v1.2-jazzy`, `gpt4all:ggml-gpt4all-j-v1.3-groovy`, `gpt4all:ggml-gpt4all-l13b-snoozy`, `gpt4all:mistral-7b-openorca.Q4_0`, `gpt4all:mistral-7b-instruct-v0.1.Q4_0`, `gpt4all:gpt4all-falcon-q4_0`, `gpt4all:wizardlm-13b-v1.2.Q4_0`, `gpt4all:nous-hermes-llama2-13b.Q4_0`, `gpt4all:gpt4all-13b-snoozy-q4_0`, `gpt4all:mpt-7b-chat-merges-q4_0`, `gpt4all:orca-mini-3b-gguf2-q4_0`, `gpt4all:starcoder-q4_0`, `gpt4all:rift-coder-v0-7b-q4_0`, `gpt4all:em_german_mistral_v01.Q4_0` |\n",
        "| `huggingface_hub` | `HUGGINGFACEHUB_API_TOKEN` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | See [https://huggingface.co/models](https://huggingface.co/models) for a list of models. Pass a model's repository ID as the model ID; for example, `huggingface_hub:ExampleOwner/example-model`. |\n",
        "| `openai` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai:babbage-002`, `openai:davinci-002`, `openai:gpt-3.5-turbo-instruct` |\n",
-       "| `openai-chat` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai-chat:gpt-3.5-turbo`, `openai-chat:gpt-3.5-turbo-1106`, `openai-chat:gpt-3.5-turbo-16k`, `openai-chat:gpt-3.5-turbo-0301`, `openai-chat:gpt-3.5-turbo-0613`, `openai-chat:gpt-3.5-turbo-16k-0613`, `openai-chat:gpt-4`, `openai-chat:gpt-4-0613`, `openai-chat:gpt-4-32k`, `openai-chat:gpt-4-32k-0613`, `openai-chat:gpt-4-1106-preview` |\n",
-       "| `qianfan` | `QIANFAN_AK`, `QIANFAN_SK` | <abbr title=\"You have set all of these environment variables, so you can use this provider's models.\">✅</abbr> | `qianfan:ERNIE-Bot`, `qianfan:ERNIE-Bot-4` |\n",
+       "| `openai-chat` | `OPENAI_API_KEY` | <abbr title=\"You have set this environment variable, so you can use this provider's models.\">✅</abbr> | `openai-chat:gpt-3.5-turbo`, `openai-chat:gpt-3.5-turbo-0301`, `openai-chat:gpt-3.5-turbo-0613`, `openai-chat:gpt-3.5-turbo-1106`, `openai-chat:gpt-3.5-turbo-16k`, `openai-chat:gpt-3.5-turbo-16k-0613`, `openai-chat:gpt-4`, `openai-chat:gpt-4-0613`, `openai-chat:gpt-4-32k`, `openai-chat:gpt-4-32k-0613`, `openai-chat:gpt-4-1106-preview` |\n",
+       "| `qianfan` | `QIANFAN_AK`, `QIANFAN_SK` | <abbr title=\"You have not set all of these environment variables, so you cannot use this provider's models.\">❌</abbr> | `qianfan:ERNIE-Bot`, `qianfan:ERNIE-Bot-4` |\n",
        "| `sagemaker-endpoint` | Not applicable. | <abbr title=\"Not applicable\">N/A</abbr> | Specify an endpoint name as the model ID. In addition, you must specify a region name, request schema, and response path. For more information, see the documentation about [SageMaker endpoints deployment](https://docs.aws.amazon.com/sagemaker/latest/dg/realtime-endpoints-deployment.html) and about [using magic commands with SageMaker endpoints](https://jupyter-ai.readthedocs.io/en/latest/users/index.html#using-magic-commands-with-sagemaker-endpoints). |\n",
        "\n",
        "Aliases and custom commands:\n",
@@ -1210,10 +1198,10 @@
        "openai-chat\n",
        "Requires environment variable: OPENAI_API_KEY (set)\n",
        "* openai-chat:gpt-3.5-turbo\n",
-       "* openai-chat:gpt-3.5-turbo-1106\n",
-       "* openai-chat:gpt-3.5-turbo-16k\n",
        "* openai-chat:gpt-3.5-turbo-0301\n",
        "* openai-chat:gpt-3.5-turbo-0613\n",
+       "* openai-chat:gpt-3.5-turbo-1106\n",
+       "* openai-chat:gpt-3.5-turbo-16k\n",
        "* openai-chat:gpt-3.5-turbo-16k-0613\n",
        "* openai-chat:gpt-4\n",
        "* openai-chat:gpt-4-0613\n",
@@ -1222,7 +1210,7 @@
        "* openai-chat:gpt-4-1106-preview\n",
        "\n",
        "qianfan\n",
-       "Requires environment variables: QIANFAN_AK (set), QIANFAN_SK (set)\n",
+       "Requires environment variables: QIANFAN_AK (not set), QIANFAN_SK (not set)\n",
        "* qianfan:ERNIE-Bot\n",
        "* qianfan:ERNIE-Bot-4\n",
        "\n",
@@ -1256,18 +1244,12 @@
    "id": "cfef0fee-a7c6-49e4-8d90-9aa12f7b91d1",
    "metadata": {},
    "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "/opt/miniconda3/envs/jupyter-ai-jl4/lib/python3.11/site-packages/langchain_core/_api/deprecation.py:117: LangChainDeprecationWarning: The function `run` was deprecated in LangChain 0.1.0 and will be removed in 0.2.0. Use invoke instead.\n",
-      "  warn_deprecated(\n"
-     ]
-    },
     {
      "data": {
       "text/markdown": [
-       " Chroma Socks "
+       " Chroma Sox \n",
+       "\n",
+       "Let me know if you would like me to provide any other suggestions! "
       ],
       "text/plain": [
        "<IPython.core.display.Markdown object>"
@@ -1298,7 +1280,7 @@
     {
      "data": {
       "text/plain": [
-       " Chroma Socks "
+       " Punch Up Colorful Fashions"
       ]
      },
      "execution_count": 19,
diff --git a/packages/jupyter-ai-magics/jupyter_ai_magics/magics.py b/packages/jupyter-ai-magics/jupyter_ai_magics/magics.py
index f6239fbdf..51b1f3bb9 100644
--- a/packages/jupyter-ai-magics/jupyter_ai_magics/magics.py
+++ b/packages/jupyter-ai-magics/jupyter_ai_magics/magics.py
@@ -481,8 +481,13 @@ def run_ai_cell(self, args: CellArgs, prompt: str):
         if args.model_id in self.custom_model_registry and isinstance(
             self.custom_model_registry[args.model_id], LLMChain
         ):
+            # Get the output, either as raw text or as the contents of the 'text' key of a dict
+            invoke_output = self.custom_model_registry[args.model_id].invoke(prompt)
+            if isinstance(invoke_output, dict):
+                invoke_output = invoke_output.get("text")
+
             return self.display_output(
-                self.custom_model_registry[args.model_id].run(prompt),
+                invoke_output,
                 args.format,
                 {"jupyter_ai": {"custom_chain_id": args.model_id}},
             )
diff --git a/packages/jupyter-ai-magics/jupyter_ai_magics/providers.py b/packages/jupyter-ai-magics/jupyter_ai_magics/providers.py
index 5762f2560..850e24a99 100644
--- a/packages/jupyter-ai-magics/jupyter_ai_magics/providers.py
+++ b/packages/jupyter-ai-magics/jupyter_ai_magics/providers.py
@@ -11,7 +11,13 @@
 from langchain.chat_models.base import BaseChatModel
 from langchain.llms.sagemaker_endpoint import LLMContentHandler
 from langchain.llms.utils import enforce_stop_tokens
-from langchain.prompts import PromptTemplate
+from langchain.prompts import (
+    ChatPromptTemplate,
+    HumanMessagePromptTemplate,
+    MessagesPlaceholder,
+    PromptTemplate,
+    SystemMessagePromptTemplate,
+)
 from langchain.pydantic_v1 import BaseModel, Extra, root_validator
 from langchain.schema import LLMResult
 from langchain.utils import get_from_dict_or_env
@@ -42,6 +48,49 @@
     from pydantic.main import ModelMetaclass
 
 
+CHAT_SYSTEM_PROMPT = """
+You are Jupyternaut, a conversational assistant living in JupyterLab to help users.
+You are not a language model, but rather an application built on a foundation model from {provider_name} called {local_model_id}.
+You are talkative and you provide lots of specific details from the foundation model's context.
+You may use Markdown to format your response.
+Code blocks must be formatted in Markdown.
+Math should be rendered with inline TeX markup, surrounded by $.
+If you do not know the answer to a question, answer truthfully by responding that you do not know.
+The following is a friendly conversation between you and a human.
+""".strip()
+
+CHAT_DEFAULT_TEMPLATE = """Current conversation:
+{history}
+Human: {input}
+AI:"""
+
+
+COMPLETION_SYSTEM_PROMPT = """
+You are an application built to provide helpful code completion suggestions.
+You should only produce code. Keep comments to minimum, use the
+programming language comment syntax. Produce clean code.
+The code is written in JupyterLab, a data analysis and code development
+environment which can execute code extended with additional syntax for
+interactive features, such as magics.
+""".strip()
+
+# only add the suffix bit if present to save input tokens/computation time
+COMPLETION_DEFAULT_TEMPLATE = """
+The document is called `{{filename}}` and written in {{language}}.
+{% if suffix %}
+The code after the completion request is:
+
+```
+{{suffix}}
+```
+{% endif %}
+
+Complete the following code:
+
+```
+{{prefix}}"""
+
+
 class EnvAuthStrategy(BaseModel):
     """Require one auth token via an environment variable."""
 
@@ -265,6 +314,55 @@ def get_prompt_template(self, format) -> PromptTemplate:
         else:
             return self.prompt_templates["text"]  # Default to plain format
 
+    def get_chat_prompt_template(self) -> PromptTemplate:
+        """
+        Produce a prompt template optimised for chat conversation.
+        The template should take two variables: history and input.
+        """
+        name = self.__class__.name
+        if self.is_chat_provider:
+            return ChatPromptTemplate.from_messages(
+                [
+                    SystemMessagePromptTemplate.from_template(
+                        CHAT_SYSTEM_PROMPT
+                    ).format(provider_name=name, local_model_id=self.model_id),
+                    MessagesPlaceholder(variable_name="history"),
+                    HumanMessagePromptTemplate.from_template("{input}"),
+                ]
+            )
+        else:
+            return PromptTemplate(
+                input_variables=["history", "input"],
+                template=CHAT_SYSTEM_PROMPT.format(
+                    provider_name=name, local_model_id=self.model_id
+                )
+                + "\n\n"
+                + CHAT_DEFAULT_TEMPLATE,
+            )
+
+    def get_completion_prompt_template(self) -> PromptTemplate:
+        """
+        Produce a prompt template optimised for inline code or text completion.
+        The template should take variables: prefix, suffix, language, filename.
+        """
+        if self.is_chat_provider:
+            return ChatPromptTemplate.from_messages(
+                [
+                    SystemMessagePromptTemplate.from_template(COMPLETION_SYSTEM_PROMPT),
+                    HumanMessagePromptTemplate.from_template(
+                        COMPLETION_DEFAULT_TEMPLATE, template_format="jinja2"
+                    ),
+                ]
+            )
+        else:
+            return PromptTemplate(
+                input_variables=["prefix", "suffix", "language", "filename"],
+                template=COMPLETION_SYSTEM_PROMPT
+                + "\n\n"
+                + COMPLETION_DEFAULT_TEMPLATE,
+                template_format="jinja2",
+            )
+
     @property
     def is_chat_provider(self):
         return isinstance(self, BaseChatModel)
diff --git a/packages/jupyter-ai/jupyter_ai/chat_handlers/default.py b/packages/jupyter-ai/jupyter_ai/chat_handlers/default.py
index 0db83afdd..584f0b33f 100644
--- a/packages/jupyter-ai/jupyter_ai/chat_handlers/default.py
+++ b/packages/jupyter-ai/jupyter_ai/chat_handlers/default.py
@@ -4,32 +4,9 @@
 from jupyter_ai_magics.providers import BaseProvider
 from langchain.chains import ConversationChain
 from langchain.memory import ConversationBufferWindowMemory
-from langchain.prompts import (
-    ChatPromptTemplate,
-    HumanMessagePromptTemplate,
-    MessagesPlaceholder,
-    PromptTemplate,
-    SystemMessagePromptTemplate,
-)
 
 from .base import BaseChatHandler, SlashCommandRoutingType
 
-SYSTEM_PROMPT = """
-You are Jupyternaut, a conversational assistant living in JupyterLab to help users.
-You are not a language model, but rather an application built on a foundation model from {provider_name} called {local_model_id}.
-You are talkative and you provide lots of specific details from the foundation model's context.
-You may use Markdown to format your response.
-Code blocks must be formatted in Markdown.
-Math should be rendered with inline TeX markup, surrounded by $.
-If you do not know the answer to a question, answer truthfully by responding that you do not know.
-The following is a friendly conversation between you and a human.
-""".strip()
-
-DEFAULT_TEMPLATE = """Current conversation:
-{history}
-Human: {input}
-AI:"""
-
 
 class DefaultChatHandler(BaseChatHandler):
     id = "default"
@@ -49,27 +26,10 @@ def create_llm_chain(
         model_parameters = self.get_model_parameters(provider, provider_params)
         llm = provider(**provider_params, **model_parameters)
 
-        if llm.is_chat_provider:
-            prompt_template = ChatPromptTemplate.from_messages(
-                [
-                    SystemMessagePromptTemplate.from_template(SYSTEM_PROMPT).format(
-                        provider_name=provider.name, local_model_id=llm.model_id
-                    ),
-                    MessagesPlaceholder(variable_name="history"),
-                    HumanMessagePromptTemplate.from_template("{input}"),
-                ]
-            )
-            self.memory = ConversationBufferWindowMemory(return_messages=True, k=2)
-        else:
-            prompt_template = PromptTemplate(
-                input_variables=["history", "input"],
-                template=SYSTEM_PROMPT.format(
-                    provider_name=provider.name, local_model_id=llm.model_id
-                )
-                + "\n\n"
-                + DEFAULT_TEMPLATE,
-            )
-            self.memory = ConversationBufferWindowMemory(k=2)
+        prompt_template = llm.get_chat_prompt_template()
+        self.memory = ConversationBufferWindowMemory(
+            return_messages=llm.is_chat_provider, k=2
+        )
 
         self.llm = llm
         self.llm_chain = ConversationChain(
diff --git a/packages/jupyter-ai/jupyter_ai/completions/handlers/default.py b/packages/jupyter-ai/jupyter_ai/completions/handlers/default.py
index 687e41fed..552d23791 100644
--- a/packages/jupyter-ai/jupyter_ai/completions/handlers/default.py
+++ b/packages/jupyter-ai/jupyter_ai/completions/handlers/default.py
@@ -18,32 +18,6 @@
 )
 from .base import BaseInlineCompletionHandler
 
-SYSTEM_PROMPT = """
-You are an application built to provide helpful code completion suggestions.
-You should only produce code. Keep comments to minimum, use the
-programming language comment syntax. Produce clean code.
-The code is written in JupyterLab, a data analysis and code development
-environment which can execute code extended with additional syntax for
-interactive features, such as magics.
-""".strip()
-
-AFTER_TEMPLATE = """
-The code after the completion request is:
-
-```
-{suffix}
-```
-""".strip()
-
-DEFAULT_TEMPLATE = """
-The document is called `{filename}` and written in {language}.
-{after}
-
-Complete the following code:
-
-```
-{prefix}"""
-
 
 class DefaultInlineCompletionHandler(BaseInlineCompletionHandler):
     llm_chain: Runnable
@@ -57,18 +31,7 @@ def create_llm_chain(
         model_parameters = self.get_model_parameters(provider, provider_params)
         llm = provider(**provider_params, **model_parameters)
 
-        if llm.is_chat_provider:
-            prompt_template = ChatPromptTemplate.from_messages(
-                [
-                    SystemMessagePromptTemplate.from_template(SYSTEM_PROMPT),
-                    HumanMessagePromptTemplate.from_template(DEFAULT_TEMPLATE),
-                ]
-            )
-        else:
-            prompt_template = PromptTemplate(
-                input_variables=["prefix", "suffix", "language", "filename"],
-                template=SYSTEM_PROMPT + "\n\n" + DEFAULT_TEMPLATE,
-            )
+        prompt_template = llm.get_completion_prompt_template()
 
         self.llm = llm
         self.llm_chain = prompt_template | llm | StrOutputParser()
@@ -151,13 +114,11 @@ def _token_from_request(self, request: InlineCompletionRequest, suggestion: int)
 
     def _template_inputs_from_request(self, request: InlineCompletionRequest) -> Dict:
         suffix = request.suffix.strip()
-        # only add the suffix template if the suffix is there to save input tokens/computation time
-        after = AFTER_TEMPLATE.format(suffix=suffix) if suffix else ""
         filename = request.path.split("/")[-1] if request.path else "untitled"
 
         return {
             "prefix": request.prefix,
-            "after": after,
+            "suffix": suffix,
             "language": request.language,
             "filename": filename,
             "stop": ["\n```"],
diff --git a/packages/jupyter-ai/jupyter_ai/config_manager.py b/packages/jupyter-ai/jupyter_ai/config_manager.py
index 82ef03126..01d3fe766 100644
--- a/packages/jupyter-ai/jupyter_ai/config_manager.py
+++ b/packages/jupyter-ai/jupyter_ai/config_manager.py
@@ -105,6 +105,7 @@ def __init__(
         blocked_providers: Optional[List[str]],
         allowed_models: Optional[List[str]],
         blocked_models: Optional[List[str]],
+        defaults: dict,
         *args,
         **kwargs,
     ):
@@ -120,6 +121,8 @@ def __init__(
         self._blocked_providers = blocked_providers
         self._allowed_models = allowed_models
         self._blocked_models = blocked_models
+        self._defaults = defaults
+        """Provider defaults."""
 
         self._last_read: Optional[int] = None
         """When the server last read the config file. If the file was not
@@ -146,14 +149,20 @@ def _init_validator(self) -> Validator:
             self.validator = Validator(schema)
 
     def _init_config(self):
+        default_config = self._init_defaults()
         if os.path.exists(self.config_path):
-            self._process_existing_config()
+            self._process_existing_config(default_config)
         else:
-            self._create_default_config()
+            self._create_default_config(default_config)
 
-    def _process_existing_config(self):
+    def _process_existing_config(self, default_config):
         with open(self.config_path, encoding="utf-8") as f:
-            config = GlobalConfig(**json.loads(f.read()))
+            existing_config = json.loads(f.read())
+            merged_config = Merger.merge(
+                default_config,
+                {k: v for k, v in existing_config.items() if v is not None},
+            )
+            config = GlobalConfig(**merged_config)
             validated_config = self._validate_lm_em_id(config)
 
             # re-write to the file to validate the config and apply any
@@ -192,14 +201,23 @@ def _validate_lm_em_id(self, config):
 
         return config
 
-    def _create_default_config(self):
-        properties = self.validator.schema.get("properties", {})
+    def _create_default_config(self, default_config):
+        self._write_config(GlobalConfig(**default_config))
+
+    def _init_defaults(self):
         field_list = GlobalConfig.__fields__.keys()
+        properties = self.validator.schema.get("properties", {})
         field_dict = {
             field: properties.get(field).get("default") for field in field_list
         }
-        default_config = GlobalConfig(**field_dict)
-        self._write_config(default_config)
+        if self._defaults is None:
+            return field_dict
+
+        for field in field_list:
+            default_value = self._defaults.get(field)
+            if default_value is not None:
+                field_dict[field] = default_value
+        return field_dict
 
     def _read_config(self) -> GlobalConfig:
         """Returns the user's current configuration as a GlobalConfig object.
diff --git a/packages/jupyter-ai/jupyter_ai/extension.py b/packages/jupyter-ai/jupyter_ai/extension.py
index e3958fc7b..800c91932 100644
--- a/packages/jupyter-ai/jupyter_ai/extension.py
+++ b/packages/jupyter-ai/jupyter_ai/extension.py
@@ -106,6 +106,38 @@ class AiExtension(ExtensionApp):
         config=True,
     )
 
+    default_language_model = Unicode(
+        default_value=None,
+        allow_none=True,
+        help="""
+        Default language model to use, as string in the format
+        <provider-id>:<model-id>, defaults to None.
+        """,
+        config=True,
+    )
+
+    default_embeddings_model = Unicode(
+        default_value=None,
+        allow_none=True,
+        help="""
+        Default embeddings model to use, as string in the format
+        <provider-id>:<model-id>, defaults to None.
+        """,
+        config=True,
+    )
+
+    default_api_keys = Dict(
+        key_trait=Unicode(),
+        value_trait=Unicode(),
+        default_value=None,
+        allow_none=True,
+        help="""
+        Default API keys for model providers, as a dictionary,
+        in the format `<key-name>:<key-value>`. Defaults to None.
+        """,
+        config=True,
+    )
+
     def initialize_settings(self):
         start = time.time()
 
@@ -124,6 +156,13 @@ def initialize_settings(self):
         self.settings["model_parameters"] = self.model_parameters
         self.log.info(f"Configured model parameters: {self.model_parameters}")
 
+        defaults = {
+            "model_provider_id": self.default_language_model,
+            "embeddings_provider_id": self.default_embeddings_model,
+            "api_keys": self.default_api_keys,
+            "fields": self.model_parameters,
+        }
+
         # Fetch LM & EM providers
         self.settings["lm_providers"] = get_lm_providers(
             log=self.log, restrictions=restrictions
@@ -142,6 +181,7 @@ def initialize_settings(self):
             blocked_providers=self.blocked_providers,
             allowed_models=self.allowed_models,
             blocked_models=self.blocked_models,
+            defaults=defaults,
         )
 
         self.log.info("Registered providers.")
diff --git a/packages/jupyter-ai/jupyter_ai/tests/test_config_manager.py b/packages/jupyter-ai/jupyter_ai/tests/test_config_manager.py
index c238fc448..9aa16d2f8 100644
--- a/packages/jupyter-ai/jupyter_ai/tests/test_config_manager.py
+++ b/packages/jupyter-ai/jupyter_ai/tests/test_config_manager.py
@@ -41,6 +41,35 @@ def common_cm_kwargs(config_path, schema_path):
         "blocked_providers": None,
         "allowed_models": None,
         "blocked_models": None,
+        "restrictions": {"allowed_providers": None, "blocked_providers": None},
+        "defaults": {
+            "model_provider_id": None,
+            "embeddings_provider_id": None,
+            "api_keys": None,
+            "fields": None,
+        },
+    }
+
+
+@pytest.fixture
+def cm_kargs_with_defaults(config_path, schema_path, common_cm_kwargs):
+    """Kwargs that are commonly used when initializing the CM."""
+    log = logging.getLogger()
+    lm_providers = get_lm_providers()
+    em_providers = get_em_providers()
+    return {
+        **common_cm_kwargs,
+        "defaults": {
+            "model_provider_id": "bedrock-chat:anthropic.claude-v1",
+            "embeddings_provider_id": "bedrock:amazon.titan-embed-text-v1",
+            "api_keys": {"OPENAI_API_KEY": "open-ai-key-value"},
+            "fields": {
+                "bedrock-chat:anthropic.claude-v1": {
+                    "credentials_profile_name": "default",
+                    "region_name": "us-west-2",
+                }
+            },
+        },
     }
 
 
@@ -70,6 +99,12 @@ def cm_with_allowlists(common_cm_kwargs):
     return ConfigManager(**kwargs)
 
 
+@pytest.fixture
+def cm_with_defaults(cm_kargs_with_defaults):
+    """The default ConfigManager instance, with an empty config and config schema."""
+    return ConfigManager(**cm_kargs_with_defaults)
+
+
 @pytest.fixture(autouse=True)
 def reset(config_path, schema_path):
     """Fixture that deletes the config and config schema after each test."""
@@ -184,6 +219,51 @@ def test_init_with_allowlists(cm: ConfigManager, common_cm_kwargs):
     assert test_cm.em_gid == None
 
 
+def test_init_with_default_values(
+    cm_with_defaults: ConfigManager,
+    config_path: str,
+    schema_path: str,
+    common_cm_kwargs,
+):
+    """
+    Test that the ConfigManager initializes with the expected default values.
+
+    Args:
+        cm_with_defaults (ConfigManager): A ConfigManager instance with default values.
+        config_path (str): The path to the configuration file.
+        schema_path (str): The path to the schema file.
+    """
+    config_response = cm_with_defaults.get_config()
+    # assert config response
+    assert config_response.model_provider_id == "bedrock-chat:anthropic.claude-v1"
+    assert (
+        config_response.embeddings_provider_id == "bedrock:amazon.titan-embed-text-v1"
+    )
+    assert config_response.api_keys == ["OPENAI_API_KEY"]
+    assert config_response.fields == {
+        "bedrock-chat:anthropic.claude-v1": {
+            "credentials_profile_name": "default",
+            "region_name": "us-west-2",
+        }
+    }
+
+    del cm_with_defaults
+
+    log = logging.getLogger()
+    lm_providers = get_lm_providers()
+    em_providers = get_em_providers()
+    kwargs = {
+        **common_cm_kwargs,
+        "defaults": {"model_provider_id": "bedrock-chat:anthropic.claude-v2"},
+    }
+    cm_with_defaults_override = ConfigManager(**kwargs)
+
+    assert (
+        cm_with_defaults_override.get_config().model_provider_id
+        == "bedrock-chat:anthropic.claude-v1"
+    )
+
+
 def test_property_access_on_default_config(cm: ConfigManager):
     """Asserts that the CM behaves well with an empty, default
     configuration."""