Not Diamond


    ### Not Diamond API documentation
    

Introduction

Model Routing Quickstart Guide

API reference

modelSelector

Not Diamond API

notdiamond-0001 API

Production

<p class="slate-p" placeholder="Type / for widgets">notdiamond-0001 automatically determines whether to send queries to GPT-3.5 or GPT-4, depending on which model is best-suited for your task. We&apos;ve trained notdiamond-0001 on <a href="https://notdiamond.ai/notdiamond-0001#footnote1" target="_blank">hundreds of thousands of data points</a> from robust, cross-domain evaluation benchmarks to maximize accuracy while drastically reducing costs and latency. When you call the notdiamond-0001 endpoint, it will return a label for either GPT-3.5 or GPT-4. You determine which version of each model you want to use and make the calls client-side with your own keys.</p><p class="slate-p" placeholder="Type / for widgets">Unlike deterministic routers, you&apos;ll notice that notdiamond-0001 doesn&apos;t route based on simple categories or domains. Instead, routing decisions are far more fine-grained. To get you started, here are some examples of prompts that get routed to either GPT-3.5 or GPT-4:</p><table class="slate-table"><colgroup><col style="width:43.625498007968126%"/><col/></colgroup><tbody class="slate-TableElement-tbody"><tr class="slate-tr"><td class="slate-td"><div class="slate-TableCellElement-content"><p class="slate-p" placeholder="Type / for widgets"><strong class="slate-bold">Sends to GPT-3.5</strong></p></div><div class="slate-TableCellElement-resizableWrapper" contenteditable="false"><div style="position:relative;user-select:auto;width:100%;height:100%;box-sizing:border-box;flex-shrink:0" ><div><div  style="position:absolute;user-select:none;width:10px;height:calc(100% + 12px);top:-12px;cursor:col-resize;right:-5px;z-index:20"></div></div></div><div class="slate-TableCellElement-handle"></div></div><div class="slate-TableCellElement-selectedCell" contenteditable="false"></div></td><td class="slate-td"><div class="slate-TableCellElement-content"><p class="slate-p" placeholder="Type / for widgets"><strong class="slate-bold">Sends to GPT-4</strong></p></div><div class="slate-TableCellElement-resizableWrapper" contenteditable="false"><div style="position:relative;user-select:auto;width:100%;height:100%;box-sizing:border-box;flex-shrink:0" ><div><div  style="position:absolute;user-select:none;width:10px;height:calc(100% + 12px);top:-12px;cursor:col-resize;right:-5px;z-index:20"></div></div></div><div class="slate-TableCellElement-handle"></div></div><div class="slate-TableCellElement-selectedCell" contenteditable="false"></div></td></tr><tr class="slate-tr"><td class="slate-td"><div class="slate-TableCellElement-content"><p class="slate-p" placeholder="Type / for widgets">What is prostate cancer?</p></div><div class="slate-TableCellElement-resizableWrapper" contenteditable="false"><div style="position:relative;user-select:auto;width:100%;height:100%;box-sizing:border-box;flex-shrink:0" ><div><div  style="position:absolute;user-select:none;width:10px;height:calc(100% + 12px);top:-12px;cursor:col-resize;right:-5px;z-index:20"></div></div></div><div class="slate-TableCellElement-handle"></div></div><div class="slate-TableCellElement-selectedCell" contenteditable="false"></div></td><td class="slate-td"><div class="slate-TableCellElement-content"><p class="slate-p" placeholder="Type / for widgets">Can you help me understand whether this prostate pathology report could indicate the presence of cancer?
CLINICAL DATA: A-H: ELEVATED PROSTATE, PROSTATE
[A] LEFT BASE… [continues]</p></div><div class="slate-TableCellElement-resizableWrapper" contenteditable="false"><div style="position:relative;user-select:auto;width:100%;height:100%;box-sizing:border-box;flex-shrink:0" ><div><div  style="position:absolute;user-select:none;width:10px;height:calc(100% + 12px);top:-12px;cursor:col-resize;right:-5px;z-index:20"></div></div></div><div class="slate-TableCellElement-handle"></div></div><div class="slate-TableCellElement-selectedCell" contenteditable="false"></div></td></tr><tr class="slate-tr"><td class="slate-td"><div class="slate-TableCellElement-content"><p class="slate-p" placeholder="Type / for widgets">What might the likely causes of a 401 error be?</p></div><div class="slate-TableCellElement-resizableWrapper" contenteditable="false"><div style="position:relative;user-select:auto;width:100%;height:100%;box-sizing:border-box;flex-shrink:0" ><div><div  style="position:absolute;user-select:none;width:10px;height:calc(100% + 12px);top:-12px;cursor:col-resize;right:-5px;z-index:20"></div></div></div><div class="slate-TableCellElement-handle"></div></div><div class="slate-TableCellElement-selectedCell" contenteditable="false"></div></td><td class="slate-td"><div class="slate-TableCellElement-content"><p class="slate-p" placeholder="Type / for widgets">This function is throwing a 401 error when it&apos;s being called. Do you see anything that could be contributing to that?</p><p class="slate-p" placeholder="Type / for widgets">def get_user_data(request, response, db: Session = Depends(get_db))… [continues]</p></div><div class="slate-TableCellElement-resizableWrapper" contenteditable="false"><div style="position:relative;user-select:auto;width:100%;height:100%;box-sizing:border-box;flex-shrink:0" ><div><div  style="position:absolute;user-select:none;width:10px;height:calc(100% + 12px);top:-12px;cursor:col-resize;right:-5px;z-index:20"></div></div></div><div class="slate-TableCellElement-handle"></div></div><div class="slate-TableCellElement-selectedCell" contenteditable="false"></div></td></tr><tr class="slate-tr"><td class="slate-td"><div class="slate-TableCellElement-content"><p class="slate-p" placeholder="Type / for widgets">What does this paragraph mean?: “As your perspective of the world increases not only is the pain it inflicts on you less but also its meaning… [continues]</p></div><div class="slate-TableCellElement-resizableWrapper" contenteditable="false"><div style="position:relative;user-select:auto;width:100%;height:100%;box-sizing:border-box;flex-shrink:0" ><div><div  style="position:absolute;user-select:none;width:10px;height:calc(100% + 12px);top:-12px;cursor:col-resize;right:-5px;z-index:20"></div></div></div><div class="slate-TableCellElement-handle"></div></div><div class="slate-TableCellElement-selectedCell" contenteditable="false"></div></td><td class="slate-td"><div class="slate-TableCellElement-content"><p class="slate-p" placeholder="Type / for widgets">Please help me complete this paragraph with a pared down description in the style of Karl One Knausgård:</p><p class="slate-p" placeholder="Type / for widgets">The broken cup lay on the table.</p></div><div class="slate-TableCellElement-resizableWrapper" contenteditable="false"><div style="position:relative;user-select:auto;width:100%;height:100%;box-sizing:border-box;flex-shrink:0" ><div><div  style="position:absolute;user-select:none;width:10px;height:calc(100% + 12px);top:-12px;cursor:col-resize;right:-5px;z-index:20"></div></div></div><div class="slate-TableCellElement-handle"></div></div><div class="slate-TableCellElement-selectedCell" contenteditable="false"></div></td></tr></tbody></table><p class="slate-p" placeholder="Type / for widgets"></p><p class="slate-p" placeholder="Type / for widgets"></p>

notdiamond-0001 automatically determines whether to send queries to GPT-3.5 or GPT-4, depending on which model is best-suited for your task. We've trained notdiamond-0001 on 

 from robust, cross-domain evaluation benchmarks to maximize accuracy while drastically reducing costs and latency. When you call the notdiamond-0001 endpoint, it will return a label for either GPT-3.5 or GPT-4. You determine which version of each model you want to use and make the calls client-side with your own keys.

Unlike deterministic routers, you'll notice that notdiamond-0001 doesn't route based on simple categories or domains. Instead, routing decisions are far more fine-grained. To get you started, here are some examples of prompts that get routed to either GPT-3.5 or GPT-4:

<p class="slate-p" placeholder="Type / for widgets">Our router is trained to optimize for cost and latency while maintaining high response quality. If your query can be adequately answered by GPT-3.5 then our router will return the corresponding label so that you can save on the cost of using GPT-4 without any degradation in quality. </p><p class="slate-p" placeholder="Type / for widgets">Integrating notdiamond-0001 is super easy. You can pass us a list of messages, just like you would for OpenAI&apos;s chat completion API, notdiamond-0001 will return the most suitable model to call and an estimate of the number of tokens your query will cost so that you don&apos;t call a model with a context length smaller than your message tokens.</p><p class="slate-p" placeholder="Type / for widgets"><strong class="slate-bold">You control</strong> which model you call, how you call it, and any rate limits you might want to impose.</p><h3 class="slate-h3">Step 1: Sign up</h3><p class="slate-p" placeholder="Type / for widgets">Head over to <a href="https://notdiamond.net" target="_blank">notdiamond-0001</a> to sign up. You only need your email address or simply sign up with your Google or Github accounts.</p><div style="text-align:left"><img src="https://theneo-prod-public.s3.amazonaws.com/images-1700668321838.png" alt="" style="width:20%"/></div><h3 class="slate-h3">Step 2: Generate a Not Diamond API key</h3><p class="slate-p" placeholder="Type / for widgets">Once you sign up, you will be greeted with an API key generation page. There you can generate an API key or delete an existing key.</p><h3 class="slate-h3">Step 3: Integration</h3><p class="slate-p" placeholder="Type / for widgets">The notdiamond-0001 /modelSelector endpoint can be called using your API key and providing a list of messages just like you would with OpenAI:</p><pre class="slate-CodeBlockElement slate-code_block"><div contenteditable="false"><div ><span aria-live="polite" aria-atomic="false" aria-relevant="additions text" ></span><div ><div ><div >Select...</div><div ><div  style="display:inline-block"><input type="text" autoCapitalize="none" autoComplete="off" autoCorrect="off" id="react-select-157-input" spellcheck="false" tabindex="0" value="" aria-autocomplete="list" style="box-sizing:content-box;width:1px;label:input;background:0;border:0;font-size:inherit;opacity:1;outline:0;padding:0;color:inherit"/><div style="position:absolute;top:0;left:0;visibility:hidden;height:0;overflow:scroll;white-space:pre"></div></div></div></div><div ><span ></span><div  aria-hidden="true"><svg height="20" width="20" viewBox="0 0 20 20" aria-hidden="true" focusable="false" ><path d="M4.516 7.548c0.436-0.446 1.043-0.481 1.576 0l3.908 3.747 3.908-3.747c0.533-0.481 1.141-0.446 1.574 0 0.436 0.445 0.408 1.197 0 1.615-0.406 0.418-4.695 4.502-4.695 4.502-0.217 0.223-0.502 0.335-0.787 0.335s-0.57-0.112-0.789-0.335c0 0-4.287-4.084-4.695-4.502s-0.436-1.17 0-1.615z"></path></svg></div></div></div></div></div><code class="language-json keep-markup drop-tokens"><div class="slate-code_line">{&quot;messages&quot;: [{</div><div class="slate-code_line">      &quot;role&quot;: &quot;assistant&quot;,</div><div class="slate-code_line">      &quot;content&quot;: &quot;How can I help you today?&quot;</div><div class="slate-code_line">    },</div><div class="slate-code_line">    {</div><div class="slate-code_line">      &quot;role&quot;: &quot;user&quot;,</div><div class="slate-code_line">      &quot;content&quot;: &quot;Can you write a function that counts from 1 to 10?&quot;</div><div class="slate-code_line">    }]</div><div class="slate-code_line">}</div></code></pre><p class="slate-p" placeholder="Type / for widgets">The received json response will look like this:</p><pre class="slate-CodeBlockElement slate-code_block"><div contenteditable="false"><div ><span aria-live="polite" aria-atomic="false" aria-relevant="additions text" ></span><div ><div ><div >Select...</div><div ><div  style="display:inline-block"><input type="text" autoCapitalize="none" autoComplete="off" autoCorrect="off" id="react-select-158-input" spellcheck="false" tabindex="0" value="" aria-autocomplete="list" style="box-sizing:content-box;width:1px;label:input;background:0;border:0;font-size:inherit;opacity:1;outline:0;padding:0;color:inherit"/><div style="position:absolute;top:0;left:0;visibility:hidden;height:0;overflow:scroll;white-space:pre"></div></div></div></div><div ><span ></span><div  aria-hidden="true"><svg height="20" width="20" viewBox="0 0 20 20" aria-hidden="true" focusable="false" ><path d="M4.516 7.548c0.436-0.446 1.043-0.481 1.576 0l3.908 3.747 3.908-3.747c0.533-0.481 1.141-0.446 1.574 0 0.436 0.445 0.408 1.197 0 1.615-0.406 0.418-4.695 4.502-4.695 4.502-0.217 0.223-0.502 0.335-0.787 0.335s-0.57-0.112-0.789-0.335c0 0-4.287-4.084-4.695-4.502s-0.436-1.17 0-1.615z"></path></svg></div></div></div></div></div><code class="language-json keep-markup drop-tokens"><div class="slate-code_line">{&quot;model&quot;: &quot;gpt-3.5&quot;, “token_estimate”: 32}</div></code></pre><p class="slate-p" placeholder="Type / for widgets">With this, you can then route your messages to the appropriate model. Below is an example of how you might handle routing using notdiamond-0001 in Python</p><pre class="slate-CodeBlockElement slate-code_block"><div contenteditable="false"><div ><span aria-live="polite" aria-atomic="false" aria-relevant="additions text" ></span><div ><div ><div >Select...</div><div ><div  style="display:inline-block"><input type="text" autoCapitalize="none" autoComplete="off" autoCorrect="off" id="react-select-159-input" spellcheck="false" tabindex="0" value="" aria-autocomplete="list" style="box-sizing:content-box;width:1px;label:input;background:0;border:0;font-size:inherit;opacity:1;outline:0;padding:0;color:inherit"/><div style="position:absolute;top:0;left:0;visibility:hidden;height:0;overflow:scroll;white-space:pre"></div></div></div></div><div ><span ></span><div  aria-hidden="true"><svg height="20" width="20" viewBox="0 0 20 20" aria-hidden="true" focusable="false" ><path d="M4.516 7.548c0.436-0.446 1.043-0.481 1.576 0l3.908 3.747 3.908-3.747c0.533-0.481 1.141-0.446 1.574 0 0.436 0.445 0.408 1.197 0 1.615-0.406 0.418-4.695 4.502-4.695 4.502-0.217 0.223-0.502 0.335-0.787 0.335s-0.57-0.112-0.789-0.335c0 0-4.287-4.084-4.695-4.502s-0.436-1.17 0-1.615z"></path></svg></div></div></div></div></div><code class="language-python keep-markup drop-tokens"><div class="slate-code_line">import json </div><div class="slate-code_line">import requests </div><div class="slate-code_line">import openai </div><div class="slate-code_line"></div><div class="slate-code_line">openai.api_key = &quot;&lt;YOUR OPENAI API KEY&gt;&quot; </div><div class="slate-code_line">nd_api_key = &quot;&lt;YOUR ND API KEY&gt;&quot;</div><div class="slate-code_line"></div><div class="slate-code_line">url = &quot;https://not-diamond-backend.onrender.com/modelSelector/&quot; </div><div class="slate-code_line"></div><div class="slate-code_line">messages = { </div><div class="slate-code_line">  &quot;messages&quot;: [</div><div class="slate-code_line">    {</div><div class="slate-code_line">      &quot;role&quot;: &quot;assistant&quot;,</div><div class="slate-code_line">      &quot;content&quot;: &quot;How can I help you today?&quot;</div><div class="slate-code_line">    },</div><div class="slate-code_line">    {</div><div class="slate-code_line">      &quot;role&quot;: &quot;user&quot;,</div><div class="slate-code_line">      &quot;content&quot;: &quot;Can you write a function that counts from 1 to 10?&quot;</div><div class="slate-code_line">    }</div><div class="slate-code_line">  ]</div><div class="slate-code_line">}</div><div class="slate-code_line"></div><div class="slate-code_line">payload = json.dumps(messages)</div><div class="slate-code_line">headers = {</div><div class="slate-code_line">  &quot;Authorization&quot;: f&quot;Bearer {nd_api_key}&quot;,</div><div class="slate-code_line">  &quot;Content-Type&quot;: &quot;application/json&quot;</div><div class="slate-code_line">}</div><div class="slate-code_line"></div><div class="slate-code_line">response = requests.request(&quot;POST&quot;, url, headers=headers, data=payload).json()</div><div class="slate-code_line"></div><div class="slate-code_line">model = response[&quot;model&quot;]</div><div class="slate-code_line">token_estimate = response[&quot;token_estimate&quot;]</div><div class="slate-code_line"></div><div class="slate-code_line">if model == &quot;gpt-3.5&quot; and token_estimate &lt;= 16384:</div><div class="slate-code_line">  openai_response = openai.ChatCompletion.create(model=&quot;gpt-3.5-turbo&quot;, messages=messages)</div><div class="slate-code_line">elif model == &quot;gpt-4&quot; and token_estimate &lt;= 8192:</div><div class="slate-code_line">  openai_response = openai.ChatCompletion.create(model=&quot;gpt-4&quot;, messages=messages)</div><div class="slate-code_line">elif model == &quot;gpt-4&quot; and token_estimate &gt; 8192:</div><div class="slate-code_line">  openai_response = openai.ChatCompletion.create(model=&quot;gpt-4-32k&quot;, messages=messages)</div><div class="slate-code_line">else:</div><div class="slate-code_line">  openai_response = openai.ChatCompletion.create(model=&quot;gpt-4-32k&quot;, messages=messages)</div><div class="slate-code_line">  </div><div class="slate-code_line">response = openai_response.choices[0].message.content</div><div class="slate-code_line">print(f&quot;This question is answered by {model}:\n{response}&quot;)</div></code></pre><div  data-type="warning" style="background-color:#FFF8E1;color:#f0b800"><div ><svg width="18" height="16" viewBox="0 0 18 16" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M0.916637 15.5009C0.791637 15.5009 0.683998 15.4732 0.59372 15.4176C0.503443 15.362 0.430526 15.2857 0.37497 15.1884C0.319415 15.0912 0.288165 14.9905 0.28122 14.8864C0.274276 14.7822 0.305526 14.6745 0.37497 14.5634L8.4583 0.605103C8.52775 0.493991 8.60761 0.41413 8.69789 0.365519C8.78816 0.316908 8.88886 0.292603 8.99997 0.292603C9.11108 0.292603 9.21178 0.316908 9.30205 0.365519C9.39233 0.41413 9.47219 0.493991 9.54164 0.605103L17.625 14.5634C17.6944 14.6745 17.7257 14.7822 17.7187 14.8864C17.7118 14.9905 17.6805 15.0912 17.625 15.1884C17.5694 15.2857 17.4965 15.362 17.4062 15.4176C17.3159 15.4732 17.2083 15.5009 17.0833 15.5009H0.916637ZM1.99997 14.2509H16L8.99997 2.1676L1.99997 14.2509ZM9.0833 13.0634C9.26386 13.0634 9.41316 13.0044 9.53122 12.8864C9.64928 12.7683 9.7083 12.619 9.7083 12.4384C9.7083 12.2579 9.64928 12.1086 9.53122 11.9905C9.41316 11.8725 9.26386 11.8134 9.0833 11.8134C8.90275 11.8134 8.75344 11.8725 8.63539 11.9905C8.51733 12.1086 8.4583 12.2579 8.4583 12.4384C8.4583 12.619 8.51733 12.7683 8.63539 12.8864C8.75344 13.0044 8.90275 13.0634 9.0833 13.0634ZM9.0833 10.7509C9.26386 10.7509 9.41316 10.6919 9.53122 10.5739C9.64928 10.4558 9.7083 10.3065 9.7083 10.1259V6.70927C9.7083 6.52871 9.64928 6.37941 9.53122 6.26135C9.41316 6.1433 9.26386 6.08427 9.0833 6.08427C8.90275 6.08427 8.75344 6.1433 8.63539 6.26135C8.51733 6.37941 8.4583 6.52871 8.4583 6.70927V10.1259C8.4583 10.3065 8.51733 10.4558 8.63539 10.5739C8.75344 10.6919 8.90275 10.7509 9.0833 10.7509Z" fill="#F1C21B"></path></svg></div><div ><p class="slate-p" placeholder="Type / for widgets"><strong class="slate-bold">Note to beta testers:</strong> Our router may take up to 10 seconds to spin up the first time you call it. Subsequent calls should be well under 1 second. Once we move out of beta there will be no initial spin up time.</p></div></div><p class="slate-p" placeholder="Type / for widgets"></p>

Our router is trained to optimize for cost and latency while maintaining high response quality. If your query can be adequately answered by GPT-3.5 then our router will return the corresponding label so that you can save on the cost of using GPT-4 without any degradation in quality. 

Integrating notdiamond-0001 is super easy. You can pass us a list of messages, just like you would for OpenAI's chat completion API, notdiamond-0001 will return the most suitable model to call and an estimate of the number of tokens your query will cost so that you don't call a model with a context length smaller than your message tokens.

 which model you call, how you call it, and any rate limits you might want to impose.

 to sign up. You only need your email address or simply sign up with your Google or Github accounts.

Once you sign up, you will be greeted with an API key generation page. There you can generate an API key or delete an existing key.

The notdiamond-0001 /modelSelector endpoint can be called using your API key and providing a list of messages just like you would with OpenAI:

The received json response will look like this:

With this, you can then route your messages to the appropriate model. Below is an example of how you might handle routing using notdiamond-0001 in Python

<p class="slate-p" placeholder="Type / for widgets">The Not Diamond API is organized around <a href="http://en.wikipedia.org/wiki/Representational_State_Transfer" target="_blank">REST</a>. Our API has predictable resource-oriented URLs, accepts <a href="https://en.wikipedia.org/wiki/POST_(HTTP)#Use_for_submitting_web_forms" target="_blank">form-encoded</a> request bodies, returns <a href="http://www.json.org/" target="_blank">JSON-encoded</a> responses, and uses standard HTTP response codes, authentication, and verbs.</p><div class="slate-hr"><hr contenteditable="false" class="slate-HrElement-hr"/></div><h3 class="slate-h3">Authentication</h3><p class="slate-p" placeholder="Type / for widgets">The Not Diamond API uses API keys for authentication. <a href="https://notdiamond.net" target="_blank">Login</a> to your account to retrieve the API key you&apos;ll use in your requests.</p><p class="slate-p" placeholder="Type / for widgets">All API requests should include your API key in an <code class="slate-code">Authorization</code> HTTP header as follows:</p><pre class="slate-CodeBlockElement slate-code_block"><div contenteditable="false"><div ><span aria-live="polite" aria-atomic="false" aria-relevant="additions text" ></span><div ><div ><div >Select...</div><div ><div  style="display:inline-block"><input type="text" autoCapitalize="none" autoComplete="off" autoCorrect="off" id="react-select-161-input" spellcheck="false" tabindex="0" value="" aria-autocomplete="list" style="box-sizing:content-box;width:1px;label:input;background:0;border:0;font-size:inherit;opacity:1;outline:0;padding:0;color:inherit"/><div style="position:absolute;top:0;left:0;visibility:hidden;height:0;overflow:scroll;white-space:pre"></div></div></div></div><div ><span ></span><div  aria-hidden="true"><svg height="20" width="20" viewBox="0 0 20 20" aria-hidden="true" focusable="false" ><path d="M4.516 7.548c0.436-0.446 1.043-0.481 1.576 0l3.908 3.747 3.908-3.747c0.533-0.481 1.141-0.446 1.574 0 0.436 0.445 0.408 1.197 0 1.615-0.406 0.418-4.695 4.502-4.695 4.502-0.217 0.223-0.502 0.335-0.787 0.335s-0.57-0.112-0.789-0.335c0 0-4.287-4.084-4.695-4.502s-0.436-1.17 0-1.615z"></path></svg></div></div></div></div></div><code class="language-shell keep-markup drop-tokens"><div class="slate-code_line">--header &apos;Authorization: Bearer &lt;ND_API_Key&gt;&apos;</div></code></pre><div  data-type="warning" style="background-color:#FFF8E1;color:#f0b800"><div ><svg width="18" height="16" viewBox="0 0 18 16" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M0.916637 15.5009C0.791637 15.5009 0.683998 15.4732 0.59372 15.4176C0.503443 15.362 0.430526 15.2857 0.37497 15.1884C0.319415 15.0912 0.288165 14.9905 0.28122 14.8864C0.274276 14.7822 0.305526 14.6745 0.37497 14.5634L8.4583 0.605103C8.52775 0.493991 8.60761 0.41413 8.69789 0.365519C8.78816 0.316908 8.88886 0.292603 8.99997 0.292603C9.11108 0.292603 9.21178 0.316908 9.30205 0.365519C9.39233 0.41413 9.47219 0.493991 9.54164 0.605103L17.625 14.5634C17.6944 14.6745 17.7257 14.7822 17.7187 14.8864C17.7118 14.9905 17.6805 15.0912 17.625 15.1884C17.5694 15.2857 17.4965 15.362 17.4062 15.4176C17.3159 15.4732 17.2083 15.5009 17.0833 15.5009H0.916637ZM1.99997 14.2509H16L8.99997 2.1676L1.99997 14.2509ZM9.0833 13.0634C9.26386 13.0634 9.41316 13.0044 9.53122 12.8864C9.64928 12.7683 9.7083 12.619 9.7083 12.4384C9.7083 12.2579 9.64928 12.1086 9.53122 11.9905C9.41316 11.8725 9.26386 11.8134 9.0833 11.8134C8.90275 11.8134 8.75344 11.8725 8.63539 11.9905C8.51733 12.1086 8.4583 12.2579 8.4583 12.4384C8.4583 12.619 8.51733 12.7683 8.63539 12.8864C8.75344 13.0044 8.90275 13.0634 9.0833 13.0634ZM9.0833 10.7509C9.26386 10.7509 9.41316 10.6919 9.53122 10.5739C9.64928 10.4558 9.7083 10.3065 9.7083 10.1259V6.70927C9.7083 6.52871 9.64928 6.37941 9.53122 6.26135C9.41316 6.1433 9.26386 6.08427 9.0833 6.08427C8.90275 6.08427 8.75344 6.1433 8.63539 6.26135C8.51733 6.37941 8.4583 6.52871 8.4583 6.70927V10.1259C8.4583 10.3065 8.51733 10.4558 8.63539 10.5739C8.75344 10.6919 8.90275 10.7509 9.0833 10.7509Z" fill="#F1C21B"></path></svg></div><div ><p class="slate-p" placeholder="Type / for widgets"><strong class="slate-bold">Remember that your API key is a secret!</strong></p><p class="slate-p" placeholder="Type / for widgets">Do not share it with others or expose it in any client-side code (browsers, apps). Production requests must be routed through your own backend server where your API key can be securely loaded from an environment variable or key management service.</p></div></div><p class="slate-p" placeholder="Type / for widgets"></p>

. Our API has predictable resource-oriented URLs, accepts 

 responses, and uses standard HTTP response codes, authentication, and verbs.

The Not Diamond API uses API keys for authentication. 

 to your account to retrieve the API key you'll use in your requests.

All API requests should include your API key in an 

<p class="slate-p" placeholder="Type / for widgets">This endpoint picks the most suitable model for your query. The response will be a json string with a label for which model to call based on the content of the query (either ‘GPT-3.5’ or ‘GPT-4’) and an estimated token count for the input. </p><p class="slate-p" placeholder="Type / for widgets"></p><p class="slate-p" placeholder="Type / for widgets"></p>

This endpoint picks the most suitable model for your query. The response will 

be a json string with a label for which model to call 

(either ‘GPT-3.5’ or ‘GPT-4’) and an estimated token count for the input.