Chorus

What it is

Chorus is the conversation intelligence product ZoomInfo acquired in 2021 and now sells as the engagement layer of its go-to-market platform. It records calls, meetings and emails, transcribes them, and indexes the result so a rep or manager can find the moment a competitor came up or a pricing objection landed without rewatching an hour of Zoom. Gong is the category leader it most resembles.

What changed in 2026 is not the recording — it is where the transcript index is reachable from. Chorus is no longer only a web app you log into. Its corpus is now queryable by AI agents through the ZoomInfo MCP server, which moves the pitch from “we record your calls” to “your call history is an addressable data source.” Two other things moved with it: chorus.ai now bounces to the ZoomInfo product page rather than serving a site of its own, and the standalone brand has dropped out of ZoomInfo’s top-level Products menu, which lists Copilot, GTM Workspace, GTM Studio and ZoomInfo MCP. Chorus sits a level down, as “Conversation intelligence” under Solutions.

Why it shows up in RevOps stacks

• The transcript corpus is exposed over MCP. The conversation_intelligence tool, shipped in the 21 July 2026 release, lets an MCP client — Claude, ChatGPT, or a custom agent — reason over indexed recordings, emails and meeting transcripts captured through Chorus or any connected third-party conferencing provider, and return a cited answer to questions like which competitors came up on an account or what your team committed to. A companion browse_engagements tool returns the meeting-and-email timeline for a company or contact. Before those two, an agent could list engagement records but had no way to read across their contents.
• The August 2026 release added skills, not just tools. ZoomInfo now distributes prebuilt workflows through its MCP skills library: Call Recap, Call Coaching, Commitment Ledger, AE-to-CS Handoff, Account Relationship Recap and Account Health. Each returns a finished artifact — a coaching read tied to specific moments in a transcript, a two-sided ledger of promises marked open or done — instead of raw records to assemble yourself.
• Methodology fields get written back automatically. Since the April 2026 release, Chorus extracts MEDDIC, MEDDPICC, BANT or SPIN values from discovery and qualification calls and writes them to Salesforce or HubSpot opportunity fields after review in a preview modal. An admin maps the framework once; reps can trigger the update from the Slack meeting recap without opening Chorus.
• Call participants resolve against the contact graph you already bought. If ZoomInfo is the data layer, the buying committee on a recording is already matched to titles and reporting lines without a separate enrichment step.

Pricing reality

Chorus has no published price. ZoomInfo’s pricing page is behind a bot wall that will not serve to automated clients, and the Chorus product page routes to Request Demo or Contact sales. Verified quote-gated 2026-08-22; unchanged from the previous check.

The more useful fact is that Chorus is not sold standalone. It is an add-on or a higher-tier unlock on a ZoomInfo platform contract, so the number to budget is the platform contract, not a per-rep line item.

• What ZoomInfo contracts actually run. Vendr’s dataset puts the median ZoomInfo contract at $33,500 per year across 1,570 purchases, with a $7,200 to $155,190 range and average savings near 22%.
• Against the leader. Vendr’s Gong median is $54,900 per year across 1,128 purchases, in an $11,195 to $204,030 range, with average savings of 14%. A conversation-intelligence-only Gong contract runs roughly 1.6× the median ZoomInfo contract that carries Chorus plus the contact database.
• Levers. ZoomInfo’s fiscal quarters end in March, June, September and December. Vendr reports 15-30% off for two-to-three-year terms, 20-35% at 50+ seats combined with a multi-year commitment, and 10-20% for annual prepayment. Onboarding and implementation fees are commonly waived on multi-year deals.
• The meter under the flat number. Data credits and AI credits are allocated, not unlimited, and overage is where these budgets break. Negotiate the credit pool up front; buying blocks mid-term is the expensive path.

Best for

RevOps and enablement leaders at companies already standardized on ZoomInfo for contact data, in the 20-200 rep band, where the buying case is “make our existing call history queryable by agents” rather than “buy the best coaching interface.” The MCP path is the single strongest reason to pick Chorus over Gong in 2026, and it only pays back for teams that genuinely run agents against GTM data.

Not for teams that do not already buy ZoomInfo for data. On coaching workflow and deal-inspection UX, Chorus is the weaker product against Gong, and taking on a ZoomInfo platform contract to obtain it inverts the cost case entirely. Below roughly 20 reps the platform contract dominates the per-rep value even for existing customers — Fathom covers recording, transcription and summaries at a per-seat price in the low tens of dollars.

Versus the alternative

• vs Gong. The category leader by share and the stronger product on deal inspection, coaching workflow and forecast rollup, and it ships a hosted MCP server of its own. Pick Gong when conversation intelligence is the primary revenue system of record and the budget supports a standalone contract at the $54,900 median. Pick Chorus when ZoomInfo is already the data layer and the incremental cost is an add-on line rather than a second platform.
• vs Clari. Post-merger Clari sells conversation intelligence inside a forecasting platform, and Clari and Salesloft shipped an MCP server in April 2026. Pick Clari when forecast accuracy is the problem the executive team is actually funding and call recordings are the evidence underneath it, not the deliverable.
• vs Fathom. The low-cost entrant that has taken the most ground at the bottom of the market: flat per-seat, bot-free capture, its own MCP server. Pick Fathom when the requirement is recordings, transcripts and summaries for a small team with no coaching program and no CRM write-back requirement.

Watch-outs

• Chorus is a feature of a platform contract, not a product you can leave behind. Churn ZoomInfo and you churn your call history. Guard: get the transcript export path and file format written into the order form before signing, and run an export quarterly so the corpus is not hostage to a renewal negotiation.
• The MCP credit treatment is not symmetric across tools. ZoomInfo states explicitly that browse_engagements consumes neither data credits nor AI credits, and makes no equivalent statement about conversation_intelligence. Guard: get the credit treatment of conversation_intelligence in writing before building agent workflows on it, then instrument actual call volume through the first month and extrapolate.
• MCP answers depend on integrations that may not be connected. browse_engagements requires at least one active calendar, email or meeting integration on the account. Guard: audit which reps have connected their calendar and mailbox before promising coverage to anyone — partial connection produces answers that are confidently incomplete rather than visibly empty.
• Automatic CRM writes from transcripts create a review burden that gets skipped. The MEDDIC and BANT extraction writes to opportunity fields only after human approval, and the approval step is exactly what a rep drops under quota pressure. Guard: keep the preview modal mandatory in the admin config and audit a monthly sample of AI-populated fields against the source recording.
• The standalone brand is being absorbed. chorus.ai redirects, the name is off the Products menu, and ZoomInfo’s own MCP tooling is named for the capability rather than the product. Guard: do not build enablement material, runbooks or integration naming on the “Chorus” brand — reference the ZoomInfo product page and the MCP tool names, which are what will survive the next repackaging.