AI Workflows and the New Plans
We are restructuring Spun's plans around the way you use AI, and we want you to know about it before anything changes. This page shows what is new in Spun's AI today, when to use Weave and when to use your own AI assistant, and the plan changes that are coming.
Spun's AI now works in two ways. AI Assistance answers when you ask: a question to Weave, a summary, a draft, a translation. AI Workflows keep working for you on a schedule: a digest that is ready every morning, a monitor that watches your groups, a reminder that goes out when someone has not checked in.
The plan changes described below are upcoming. They take effect on [DATE]. Nothing changes on your account before then, and nothing you use today is taken away.
What is new
Chat digests
A chat digest answers a standing question about your chats and keeps the answer current as messages arrive: "What is still open in the partner groups?", "Any complaints this week?", or simply "Summarize".
- One chat, chosen chats or all chats. A digest covers one chat, up to 25 chats and groups you pick, or every chat in your workspace, including chats that start later.
- Evidence you can check. Every claim in the answer cites the exact messages it came from, and the digest shows how many messages it analyzed.
- On a schedule. Refresh on demand or every day, and add any digest as a section of your morning Daily Summary.
- Cost shown before you create it. The app shows an estimate of what the digest will use each month before you save it.
- Voice notes included. Digests name voice notes and their length, and note when a transcript is not ready yet.
- Version history. Every refresh is kept, and you can roll back.
Group monitors
Group Intelligence monitors watch the WhatsApp groups you choose and raise a signal when something matters: a request, a lead, a complaint, or a topic you named. Each signal carries its confidence and a status you can move through reviewing, assigned, contacted or dismissed, with an AI draft reply when you want one. Alerts arrive immediately, as a daily digest or weekly.
Staff response-time monitors are new. Pick groups and a time limit, and a signal appears when a customer's request in one of those groups gets no reply from your staff within that time. These monitors use no AI and spend no tokens.
Staff and group reports
Ask Weave how your team is doing and it reads one report covering up to 92 days: messages per team member, the groups each one is active in, their median response time, and the share of requests answered within an hour. Ask about your groups and it reads a group report: the busiest groups, the requests still waiting, the most active members and the median response time per group. The group chats guide shows how to tell Weave who your team is and which questions work best.
Check-in reminders
Ask Weave to remind the people in a group who have not sent a voice note, or any message, by a set time. Weave explains the rule and offers to create it. Every reminder waits in Approvals for a person to send; nothing goes out on its own. See Check-in Rules.
Connected AI: ChatGPT, Claude and other assistants
With Connected AI, the assistant you already use reads your Spun workspace in its own chat window.
- Who wrote each group message. Every inbound group message carries the sender's number, your contact name for that number, and whether WhatsApp hid the number. Hidden-number members get a stable key, so the assistant can still count and group their messages; Spun never fills in a number it does not have.
- Digests from your assistant. It can list your digests, read a digest's prepared answer in one call however many messages it covers, and, with the right permission, create or edit a digest.
- Sign in with Spun. Assistants such as Claude connect through a consent screen with your Spun account; other clients use a key with one of three presets, from read-only to assistant with sends.
- Your usage, visible. Settings shows the month's connected-AI calls against your plan, and the assistant can check its own usage.
Weave or your own assistant: which to use
Both read the same workspace with the same permissions. They differ in where they run, what they cost you, and what each is good at.
| Weave, inside Spun | Your own assistant through Connected AI | |
|---|---|---|
| Where | The Weave tab in Spun, on desktop and phone, with voice mode | ChatGPT, Claude or another assistant, in its own window |
| What it costs you | Each question uses AI credit, because the AI that answers runs on Spun's account | Reads use no AI credit from [DATE] (today each read counts 2 tokens inside your allowance), and a draft your assistant writes itself uses none either: its own subscription pays for its thinking. Credit is used only when you ask Spun's AI to do the work, for example to generate a draft with Spun's tone settings or to rebuild a digest |
| Numbers across your whole history: counts, rankings, trends, "which customers went quiet" | One step. Weave queries the workspace directly, so a question such as "messages per label per week" or "groups with the most unanswered requests" comes back as a table | Works from a fixed set of tools and reads messages page by page, so exact numbers across thousands of messages take many calls and may run into the assistant's own limits |
| Staff and group reports, digests, message statistics | Yes | Yes: the same reports and digests are available to the assistant, and reading a prepared digest is one call |
| Setting up workflows | Ask Weave to create a digest, a monitor or a check-in rule in context; it shows a confirmation first | Can create and edit digests. Monitors and check-in rules are set up in Spun or by asking Weave. All scheduled work runs in Spun either way |
| Acting in Spun | Proposes labels, tasks, alarms, opening a chat, importing contacts or linking a hidden member, each with a confirm button | Can create tasks, labels and drafts, and propose sends, polls, locations, status posts and group changes, all through the approval queue |
| Putting Spun next to the rest of your work | Sees Spun only | Combines your WhatsApp data with your files, email, calendar and the web, in the assistant you already use |
| Speed | Answers inside Spun with direct access, so multi-step questions finish in fewer rounds | Each tool call is a round trip from the assistant to Spun |
| Help with Spun itself | Answers from Spun's help articles | Can search the same help articles |
A simple rule: use your own assistant for everyday reading, drafting in your own style, and anything that mixes Spun with your other tools, because it costs you nothing in AI credit. Use Weave when you want exact numbers over a lot of history, a report or digest set up from a conversation, or an action taken in Spun with one confirmation.
The three tiers
| Trial, 7 days | Starter | Pro | Max | |
|---|---|---|---|---|
| Price per month | free | $49 | $100 | $200 |
| Positioning | a full preview of Max | AI Assistance | AI Workflows. Recommended for daily business operations | Advanced AI Workflows |
| AI credit included | $2.50 for the week | $5 a month | $15 a month | $40 a month |
| Ask Weave: questions, summaries, drafts, translation, voice notes, manual digests, confirmed actions | yes | yes | yes | yes |
| Workflows on a schedule: digests, monitors, check-in reminders | all, for the week | 1 | up to 5 | unlimited |
| A digest or monitor can cover | one chat, up to 25 chats and groups, or all chats | same | same | same |
| Staff reporting | everything in Max | not included | response times and unanswered requests | team comparisons, group comparisons, and staff response-time monitors with your own time limit |
| Connected AI (ChatGPT, Claude and others) | 300 reads for the week | 2,000 reads a month, then the Connected AI add-on at $25 a month | included | included |
| Business automation: chatflows, webhooks, general REST API and integration credentials, Google Sheets sync | yes | not included (keys for connected AI assistants stay available) | yes | yes |
| Shared inbox with your team | not included (a trial can join a team's shared inbox) | yes | yes | yes |
| Ordinary and scheduled messaging, broadcasts | yes, never charged against AI credit | yes | yes | yes |
| Message history | 30 days | 90 days | 1 year | unlimited |
| Storage | 5 GB | 10 GB | 100 GB | 500 GB |
| Voice note transcription, maximum length | 5 min | 5 min | 10 min | 10 min |
| Extra WhatsApp number | $49 a month | $29 a month | $29 a month |
A workflow is one digest, monitor or check-in rule that runs on a schedule. A paused workflow does not count. Workflows use your AI credit when they run; a staff response-time monitor uses none.
The upcoming plan restructure
These are the changes we are planning for [DATE]. We are sharing them now so you are aware of the restructure before it happens.
New names, same prices. Pro at $49 becomes Starter. Max 10x at $100 becomes Pro. Max 40x at $200 becomes Max. Prices do not change.
AI credit instead of a token count. Each plan includes a fixed amount of AI credit: $5 on Starter, $15 on Pro, $40 on Max. Today's allowances are 100,000, 1,000,000 and 4,000,000 tokens; the new ones are 1,000,000, 3,000,000 and 8,000,000 tokens, where 1,000,000 tokens is $5 of credit. AI usage is charged against that credit at the AI provider's published rates plus 30 percent, and you can see the breakdown by feature in Settings → Plan & Usage.
Weave is charged at its real rate. Until now a Weave question was charged well below what the AI behind it costs. From [DATE] it is charged at the real rate, so a question draws more tokens than it does today. The larger allowances are designed to absorb this for normal use, and every account that has used Weave receives a one-time credit of 200,000 tokens.
Extra credit at face value. When you need more, credit is sold at $5 per 1,000,000 tokens in packs of $5, $10, $25 and $50, under a monthly spending cap you set. The current packs at higher prices are withdrawn. Anything you have already bought keeps its full buying power.
Connected AI is included in Pro and Max. Starter includes 2,000 connected-AI reads a month and can add unlimited reads for $25 a month. Pro and Max include unlimited reads within the usual rate limits. Connected-AI reads no longer draw from your AI credit.
One meter for digests. Digests are measured in AI credit only. The separate monthly allowance of digest messages goes away.
Workflows by plan. Starter includes one scheduled workflow, Pro up to five, Max unlimited. Business automation (chatflows, webhooks, general REST API and integration credentials, and Google Sheets sync) is part of Pro and Max for new subscriptions; Starter keeps the keys and sign-in used by connected AI assistants.
The trial shows everything. New trials get 7 days of Max with $2.50 of AI credit and 300 connected-AI reads. When the trial ends, workflows beyond the chosen plan pause with their settings kept, and an upgrade resumes them.
If you are already a customer
- Your price stays the same, and you keep everything you have today, including any automation you set up.
- You receive the larger AI credit of your plan from [DATE], and the one-time 200,000-token credit if your account has used Weave.
- Credit you bought keeps its buying power.
- Your subscription price does not change. What changes is how AI usage is deducted from your credit, the size of the credit included in your plan, and the price of extra credit, as described above.
What does not change
- Sending and receiving messages, broadcasts and scheduled sends never draw from AI credit, on any plan.
- Sharing your inbox with your team is part of every paid plan.
- Every send proposed by Weave or by a connected assistant waits for a person to approve it, unless your organization has chosen otherwise in its settings.
- The limits that protect your WhatsApp number are the same on every plan.