Updates to Workers AIhttps://developers.cloudflare.com/workers-ai/changeloghttps://developers.cloudflare.com/workers-ai/changelog/#workers-ai-birthday-week-2024-announcementshttps://developers.cloudflare.com/workers-ai/changelog/#workers-ai-birthday-week-2024-announcements<ul> <li>Meta Llama 3.2 1B, 3B, and 11B vision is now available on Workers AI</li> <li><code>@cf/black-forest-labs/flux-1-schnell</code> is now available on Workers AI</li> <li>Workers AI is fast! Powered by new GPUs and optimizations, you can expect faster inference on Llama 3.1, Llama 3.2, and FLUX models.</li> <li>No more neurons. Workers AI is moving towards <a href="https://developers.cloudflare.com/workers-ai/platform/pricing">unit-based pricing</a></li> <li>Model pages get a refresh with better documentation on parameters, pricing, and model capabilities</li> <li>Closed beta for our Run Any* Model feature, <a href="https://forms.gle/h7FcaTF4Zo5dzNb68">sign up here</a></li> <li>Check out the <a href="https://blog.cloudflare.com/workers-ai">product announcements blog post</a> for more information</li> <li>And the <a href="https://blog.cloudflare.com/workers-ai/making-workers-ai-faster">technical blog post</a> if you want to learn about how we made Workers AI fast</li> </ul> Thu, 26 Sep 2024 00:00:00 GMThttps://developers.cloudflare.com/workers-ai/changelog/#meta-llama-31-now-available-on-workers-aihttps://developers.cloudflare.com/workers-ai/changelog/#meta-llama-31-now-available-on-workers-ai<p>Workers AI now suppoorts <a href="https://developers.cloudflare.com/workers-ai/models/llama-3.1-8b-instruct/">Meta Llama 3.1</a>.</p> Tue, 23 Jul 2024 00:00:00 GMThttps://developers.cloudflare.com/workers-ai/changelog/#new-community-contributed-tutorialhttps://developers.cloudflare.com/workers-ai/changelog/#new-community-contributed-tutorial<ul> <li>Added community contributed tutorial on how to <a href="https://developers.cloudflare.com/developer-spotlight/tutorials/creating-a-recommendation-api/">create APIs to recommend products on e-commerce sites using Workers AI and Stripe</a>.</li> </ul> Thu, 11 Jul 2024 00:00:00 GMThttps://developers.cloudflare.com/workers-ai/changelog/#introducing-embedded-function-callinghttps://developers.cloudflare.com/workers-ai/changelog/#introducing-embedded-function-calling<ul> <li>A new way to do function calling with <a href="https://developers.cloudflare.com/workers-ai/function-calling/embedded">Embedded function calling</a></li> <li>Published new <a href="https://www.npmjs.com/package/@cloudflare/ai-utils"><code>@cloudflare/ai-utils</code></a> npm package</li> <li>Open-sourced <a href="https://github.com/cloudflare/ai-utils"><code>ai-utils on Github</code></a></li> </ul> Thu, 27 Jun 2024 00:00:00 GMThttps://developers.cloudflare.com/workers-ai/changelog/#added-support-for-traditional-function-callinghttps://developers.cloudflare.com/workers-ai/changelog/#added-support-for-traditional-function-calling<ul> <li><a href="https://developers.cloudflare.com/workers-ai/function-calling/">Function calling</a> is now supported on enabled models</li> <li>Properties added on <a href="https://developers.cloudflare.com/workers-ai/models/">models</a> page to show which models support function calling</li> </ul> Wed, 19 Jun 2024 00:00:00 GMThttps://developers.cloudflare.com/workers-ai/changelog/#native-support-for-ai-gatewayshttps://developers.cloudflare.com/workers-ai/changelog/#native-support-for-ai-gateways<p>Workers AI now natively supports <a href="https://developers.cloudflare.com/ai-gateway/providers/workersai/#worker">AI Gateway</a>.</p> Tue, 18 Jun 2024 00:00:00 GMThttps://developers.cloudflare.com/workers-ai/changelog/#deprecation-announcement-for-cfmetallama-2-7b-chat-int8https://developers.cloudflare.com/workers-ai/changelog/#deprecation-announcement-for-cfmetallama-2-7b-chat-int8<p>We will be deprecating <code>@cf/meta/llama-2-7b-chat-int8</code> on 2024-06-30.</p> <p>Replace the model ID in your code with a new model of your choice:</p> <ul> <li><a href="https://developers.cloudflare.com/workers-ai/models/llama-3-8b-instruct/"><code>@cf/meta/llama-3-8b-instruct</code></a> is the newest model in the Llama family (and is currently free for a limited time on Workers AI).</li> <li><a href="https://developers.cloudflare.com/workers-ai/models/llama-3-8b-instruct-awq/"><code>@cf/meta/llama-3-8b-instruct-awq</code></a> is the new Llama 3 in a similar precision to your currently selected model. This model is also currently free for a limited time.</li> </ul> <p>If you do not switch to a different model by June 30th, we will automatically start returning inference from <code>@cf/meta/llama-3-8b-instruct-awq</code>.</p> Tue, 11 Jun 2024 00:00:00 GMThttps://developers.cloudflare.com/workers-ai/changelog/#add-new-public-loras-and-note-on-lora-routinghttps://developers.cloudflare.com/workers-ai/changelog/#add-new-public-loras-and-note-on-lora-routing<ul> <li>Added documentation on <a href="https://developers.cloudflare.com/workers-ai/fine-tunes/public-loras/">new public LoRAs</a>.</li> <li>Noted that you can now run LoRA inference with the base model rather than explicitly calling the <code>-lora</code> version</li> </ul> Wed, 29 May 2024 00:00:00 GMThttps://developers.cloudflare.com/workers-ai/changelog/#add-openai-compatible-api-endpointshttps://developers.cloudflare.com/workers-ai/changelog/#add-openai-compatible-api-endpoints<p>Added OpenAI compatible API endpoints for <code>/v1/chat/completions</code> and <code>/v1/embeddings</code>. For more details, refer to <a href="https://developers.cloudflare.com/workers-ai/configuration/open-ai-compatibility/">Configurations</a>.</p> Fri, 17 May 2024 00:00:00 GMThttps://developers.cloudflare.com/workers-ai/changelog/#add-ai-native-bindinghttps://developers.cloudflare.com/workers-ai/changelog/#add-ai-native-binding<ul> <li>Added new AI native binding, you can now run models with <code>const resp = await env.AI.run(modelName, inputs)</code></li> <li>Deprecated <code>@cloudflare/ai</code> npm package. While existing solutions using the @cloudflare/ai package will continue to work, no new Workers AI features will be supported. Moving to native AI bindings is highly recommended</li> </ul> Thu, 11 Apr 2024 00:00:00 GMT