# OpenAI models, working on your real Salesforce org. > SyncOnAI 360 can run its chat and specialist agents on OpenAI models, chosen separately for each. OpenAI also powers meeting transcription into pages and image generation in the chat, which need an OpenAI key in the workspace on every plan for now. OpenAI does not train on API traffic, and stored keys are encrypted. Source: https://synconai360.com/connectors/openai ## Key facts - **2**: Separate model settings: chat and agents - **2**: Features that use OpenAI on every plan - **0**: Training on API traffic - **1**: Key per workspace, tested before use ## Why teams want model choice on Salesforce Different teams standardise on different AI providers, and some work suits one model better than another. A Salesforce tool that locks you to one provider, or leaves model use ungoverned, makes that harder. - **Provider lock-in.** A tool tied to one model provider overrides the company's own choice. - **Scattered keys.** Each person uses their own key in their own tool, with no workspace control. - **Meetings unrecorded.** Discovery calls and design decisions are lost because nobody wrote them down. - **No guardrails.** A model's output reaches the org by copy and paste, with no checks. ## How do you run SyncOnAI 360 on OpenAI models? Under Settings, Models, choose the model the chat uses and, separately, the model the agents use, from OpenAI and Anthropic. Chat and agents can run on different providers. Whichever model answers, it reads the synced org through tools, shows the tools it used, and turns anything it drafts into a proposal with named checks, validation against the org and approval for production. - Chat and agents set separately - OpenAI or Anthropic - Same tools and guardrails either way ## How are meetings transcribed into Salesforce documentation? Press the microphone beside New page, choose the browser tab with the call, and its audio is transcribed as you go using OpenAI's speech model. When you stop, the transcript is saved as a page beside the org and project it concerns. Pages are searchable from the Knowledge Hub and the chat, so decisions from a discovery call can be cited when the work is planned. - Any call in a browser tab - Saved as a page - Searchable and cited later ## What else uses OpenAI? Image generation in the chat runs on OpenAI's image model. Together with meeting transcription, it needs an OpenAI key in the workspace on every plan for now, even where other AI is included. The settings page says which features need which key, so nobody discovers a missing key in the middle of a client call. - Image generation in chat - OpenAI key needed for both, for now - Stated in settings ## How is an OpenAI key stored and checked? Paste the key under Settings, API keys; one key per workspace. Test checks that it works before anyone relies on it. Keys are encrypted under a key held in a cloud key management service and never shown in full after saving. On the Free plan your key powers all AI, billed by OpenAI. Paid plans include AI for the chat and agents. - One key per workspace - Tested on save - Encrypted, never shown in full ## Does OpenAI train on your Salesforce data? No. OpenAI does not train on API traffic, whichever account a request uses. SyncOnAI 360 sends the context a request needs and never stores your CRM records. OpenAI appears on the subprocessors page with what it receives and where. - No training on API traffic - Records never stored - Listed as a subprocessor ## How do OpenAI models follow your org's conventions? Org rules set naming conventions, forbidden patterns and the preferred automation type for each org, and anything the model drafts is built inside them. Before building, the chat checks what already exists, so it extends what is there rather than adding a second way of doing the same thing. Specialist agents add their own instructions and skills, and their tools can be allowed, denied or held for confirmation, whichever model runs them. - Org rules the model must follow - Existing solutions checked first - Agent tools under your control ## Using OpenAI models 1. **Choose.** Set the chat and agent models under Settings. 2. **Add a key.** Needed on Free, and for transcription and images. 3. **Test.** Check the key works before you rely on it. 4. **Work.** Ask, build, transcribe, and ship through proposals. ## OpenAI on Salesforce, loose versus governed | Without | With SyncOnAI 360 | |---|---| | One provider imposed by the tool | OpenAI or Anthropic, per setting | | Keys in every person's tools | One encrypted workspace key | | Meetings lost after the call | Transcripts saved as pages | | Output pasted into the org | Drafts become checked proposals | ## Who uses OpenAI models - **Consultancies.** Transcribe discovery calls into pages beside the client's org, and keep requirements searchable. - **Platform teams.** Run the workspace on the provider the company has already approved. - **Delivery managers.** Turn meeting decisions into documented work without a separate note-taking tool. - **IT and platform leaders.** Control which keys are used and see that no provider trains on the traffic. ## Frequently asked questions ### Can SyncOnAI 360 run on OpenAI models? Yes. The chat and the agents can each be set to an OpenAI model under Settings, Models. ### Why do we need an OpenAI key for transcription? Meeting transcription runs on OpenAI's speech model, so it needs an OpenAI key in the workspace on every plan for now. ### Does image generation need a key too? Yes. Image generation in the chat runs on OpenAI's image model and needs an OpenAI key for now. ### Does OpenAI train on our data? No. OpenAI does not train on API traffic. ### How is the key stored? Encrypted under a key held in a cloud key management service, and never shown in full after saving. ### Can we mix providers? Yes. The chat and the agents can run on different providers. ### Do paid plans need a key? Not for the chat and agents, where AI is included. Transcription and image generation still need an OpenAI key for now. ### Can the model change our org directly? No. Anything it drafts becomes a proposal with checks, validation and approval. ### Which plans include AI? Pro, Team and Business include AI, so no key is needed. Free runs on your own Anthropic or OpenAI key, and Enterprise can include AI or use your own provider agreement. ### Can we switch provider later? Yes. Change the model setting at any time; your orgs, history and proposals stay as they are. ### Where does a meeting transcript go? It is saved as a page in the workspace, searchable from the Knowledge Hub and the chat. ### Can agents and chat use OpenAI on the Free plan? Yes, on your own OpenAI key, billed by OpenAI. The same key also covers transcription and image generation.