Click Chat in the left-hand sidebar to open it. Not there? It just hasn't been enabled yet, reach out to [email protected] and we'll turn it on for your account.

(Chat may become a paid add-on down the line.)

The first time you open Chat, the center panel shows four quick-start categories:

Click any category to see a dropdown of ready-to-run prompts. For example, clicking Daily workflows shows:

Click any prompt in that list and it runs immediately, no setup required.

If you don't see what you're after in these four categories, click Show all prompts to open the full Prompt Library instead.

Click Show all prompts (below the four quick-start categories) to open the Prompt Library, your central workspace for browsing, running, organizing, and creating reusable AI workflows.

Inside the Prompt Library, you can:

Segment-specific prompts are available directly in the Prompt Library, under categories like:

Segment Insights

Outreach prep (segment-aware)

These prompts pull directly from your Segmentation data, comparing win rate, deal size, coverage, and growth signals across your actual segments, rather than requiring you to manually cross-reference that data yourself. Just click Run to execute one directly, or Copy to adapt it first.

At the top of the Prompt Library, you'll find two tabs:

Curated: prompts shared across your entire workspace, visible to everyone on your Ocean.io account. Each prompt shows its category (e.g. Daily workflows, Account research, Build lead lists) as a colored tag above the prompt text. Good for standardizing team-wide workflows and outbound motions.

Private: prompts visible only to you. Good for personal experiments and one-off searches.

Every prompt, curated or private, draws directly on your own Ocean.io data (segmentations, leads, CRM data, closed-won deals, existing accounts, connected integrations) when it runs, so results stay grounded in your business rather than generic.

Click New Prompt in the top-right corner to open the Save prompt window. Here you can fill in:

Once filled in, click Save prompt to add it to your library.

Every prompt, curated/shared or private, draws directly on your own Ocean.io data (segmentations, leads, CRM data, closed-won deals, existing accounts, connected integrations) when it runs, so results stay grounded in your business rather than generic.

To start a new conversation, simply type your question or request into the main input box (the large "How can I help today?" field in the center of the screen) and send it, Chat takes it from there.

If you're already inside a conversation and want to start something new, click the back arrow in the top-left corner to return to the main screen, then type your new request into the input box there. The New chat icon in the left sidebar does the same thing, if you prefer using that instead.

Within an existing conversation, Chat remembers your previous filters and context, so you can just keep typing follow-ups without starting over:

"Find SaaS companies in DACH with 100–500 employees" → "Only untouched accounts" → "Now expand to Western Europe"

Each follow-up narrows or broadens the existing search rather than starting fresh.

A quick note on scope: Chat is built specifically for sales prospecting and go-to-market tasks, finding companies, identifying decision-makers, discovering lookalikes, and building prospect lists. If you ask it something outside that scope, it'll let you know and steer you back toward what it can actually help with.

AI Context is not required to use Chat, Chat works perfectly well without it. Setting it up just makes results significantly more accurate, since Chat then knows your ICP, your products, and your business context when running searches. Think of it as optional, high-value setup, not a prerequisite.

You can access AI Context two ways: through Profile → Settings → Context (under the AI section in the left settings sidebar), or quickly from within a chat by clicking the sliders icon next to Add context in the message input bar.

Enter your company website, then click Draft context. Ocean.io analyzes your site and automatically drafts a full profile of your business, what you sell, your ideal customer profile, your value proposition, your differentiators, and your competitors. This usually takes a few minutes to generate.

Don't have a website? Click I don't have a website to fill this in manually instead.

Once drafted, you'll see your context organized into:

Selecting a product shows its full profile, including:

At the company level, you'll also see:

Review what Ocean.io drafted, edit any field directly, and click Approve and save to confirm it, or Discard to throw it out and start over.

This isn't a one-time setup. You can come back and refine your AI Context anytime, add new products, update your ICP, adjust your competitors, or sharpen your value proposition as your business evolves. The more complete and accurate this is, the better Chat gets at finding the right prospects, sharpening outreach, and answering your questions.

Click the model selector on the left side of the message input bar to pick which AI powers your conversation. You'll see a few options ranging from fastest/cheapest to deepest reasoning/most expensive, the cost tier is shown next to each one. More $ signs means more credits consumed per response.

Rule of thumb: match the model to the complexity of the task, not the other way around. A quick lookup doesn't need your most expensive option, save the $$$ tier for deep account analysis or advanced ICP work, where reasoning quality actually matters more than speed.

💡 Pricing for AI usage is at cost during Beta and may change at launch.

Prefer to use your own API key instead of one of these models? See Bring Your Own API Key below.

Ocean.io supports bringing your own API key from either Anthropic or OpenAI to power your chat sessions, instead of using one of Ocean's built-in models.

To connect a key, click the sliders icon in the message input bar (next to Add context), this opens a menu with two options: Add Anthropic key and Add OpenAI key. Select the one you want and enter your key.

You can also connect a key via Profile → Settings → Providers (under the AI section in the left settings sidebar).

Once connected:

Useful for heavy AI users, teams with existing Anthropic or OpenAI billing, or anyone who wants more control over their AI infrastructure costs.

Every prompt uses credits, covering AI usage, search execution, result reveals, CSV exports, and CRM/webhook pushes.

The credit cost of each response appears directly below that response, once the request has completed. There's currently no upfront estimate shown before you run a prompt, so broad searches, complex requests, or higher-cost models can end up using more credits than expected.

Records you've already revealed generally aren't charged reveal credits again if you encounter them in a later search.

Once you have results you want to keep, you can export them to:

For larger exports, Chat states the record count and credit cost upfront in its response and waits for you to confirm before actually running the export.

The left sidebar lists your past conversations, grouped by recency (Today, Previous 7 days, Older). Use the Search conversations bar at the top of this list to find a past chat by keyword.

From the sidebar list, hover over any conversation and click the three-dot menu (...) that appears to:

Inside an open conversation, click the three-dot menu (...) in the top-right corner to:

Manual controls inside a conversation: the results panel within a chat works the same as the general Ocean.io platform UI, you're not limited to only what the AI does automatically. From here you can:

So if you want more manual control at any point, you can drop into that view and drive it yourself rather than staying purely conversational.

Chat is in beta, so your feedback matters. For bugs, incorrect results, or workflow suggestions, email [email protected], you can also share the actual chat session directly with our team to help us debug faster.