Deltas
Also written: Event deltas, differential updates, change deltas
Small updates showing what has changed since the last piece of information.
Deltas are small, incremental updates. Instead of sending a whole new file or message every time something changes, only the changes themselves are sent. This makes things faster and uses less data.
Think about a live chat. When someone types a new sentence, the chat program doesn't resend the entire conversation. It just sends the new sentence. That new sentence is a delta.
Deltas represent the incremental changes or differences between successive states of data. In AI, this is often used in streaming applications where sending the complete state repeatedly would be inefficient. For example, in speech-to-text, deltas can be small chunks of transcribed text as they become available, rather than waiting for the entire utterance to be processed.
Recent frontier AI developments focus on real-time applications like live translation and transcription. These applications rely on efficiently sending small, incremental updates, or deltas, to provide low-latency streaming experiences. Companies are optimizing their models and infrastructure to deliver these rapid, partial updates.
- GPT-Realtime-Translate ModelOpenAI · Model Documentation ·
- GPT-Live-Transcribe ModelOpenAI · Model Documentation ·
- GPT-Realtime-Whisper ModelOpenAI · Model Documentation ·
- StreamingxAI · xAI Docs: Guides ·
- Anthropic Cookbook: feat(managed_agents): add subagent live-streaming cookbookAnthropic · Anthropic Cookbook ·
- OpenAI ChatGPT
- OpenAI's real-time models like GPT-Realtime-Translate and GPT-Live-Transcribe likely use deltas to stream audio and text updates for low-latency performance.
- xAI Grok
- xAI's Grok supports streaming text outputs, indicating it can send deltas for faster, real-time responses.
- Full State
- Sending the entire current state instead of just the changes.
- Batches
- Sending data in groups rather than as small, continuous updates.
- Diffs
- A general term for differences between files; deltas are often used in a
The term 'delta' originates from the Greek letter delta (Δ), commonly used in mathematics and science to denote change or difference. In computing, it has long been used for file comparison and update mechanisms.
Use a real-time transcription service or a chatbot that supports streaming responses. Observe how the text appears word by word or sentence by sentence, rather than all at once. This is an example of deltas being sent to your device.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- Why are deltas important for AI?
- Deltas enable real-time and low-latency applications by only sending necessary updates, making services like live translation and chatbots feel more responsive.
- Is a delta the same as a full message?
- No, a delta is only the part that has changed or is new, while a full message or state contains all the information up to that point.
First seen on Subvolts Jul 22, 2026. Updated Oct 5, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.