Ada

Summary

Ada is an enterprise customer-service platform whose AI agents resolve support inquiries over voice, email, chat and social messaging. Its Reasoning Engine plans each step and runs API, Python and MCP tools inside the customer's own systems to complete the task.

Best for: Enterprises that want a long-running, well-capitalized AI support agent vendor with a public V2 REST API and an MCP server, rather than a sales-only integration surface.
Not for: Teams that want transparent public pricing, a self-serve trial to test before buying, or that have been burned before by a wide gap between polished admin tooling and the actual end-customer chat experience.

Captured 2026-09-08 · theagentsindex.com

Price: We also offer a resolution-based pricing model for enterprises with specific needs, but most customers choose conversation-based pricing for its predictability, transparency, and lower total cost of ownership.no public pricing
Free tier: NoNo free plan and no self-serve trial is published. The only documented route in is a booked expert review and custom demo; dashboard access requires an Ada-provisioned instance and a business email address.

Visit Ada Compare Find alternatives

What it does

Writes code: PartlyAda does not author application code, but it drafts sandboxed Python code tools: you describe the tool in plain language and a coding agent writes it.Connect an MCP server and its tools are discovered. Describe a code tool and a coding agent drafts it.ada.cx/platform/ada-computer
Debugs and fixes: NoAda diagnoses its own agent's conversations, not software failures; its only code surface is a small data-shaping Python tool.A code tool runs a small piece of Python to clean up, transform, or compute data before your AI Agent uses it.docs.ada.cx/docs/automation/tools/code-…
Writes tests: NoTest cases here are multi-turn conversation scenarios scored pass/fail against the AI agent's replies, not software tests.Create test cases from production conversations, trigger test runs, and review pass/fail results and evaluation rationale, all without leaving your…docs.ada.cx/mcp/introduction/overview
Reviews code: NoThe only review surface is the agent's own configuration; nothing reads a diff or a pull request.Audit your setup. "Summarize our configuration" returns every playbook, API tool, coaching entry, and custom instruction in one view.docs.ada.cx/mcp/introduction/overview
Whole-codebase work: NoThere is no repository concept: code exists only as single sandboxed snippets in a restricted Python subset.Language: Python only, a restricted subset, not full Python. A few standard-library modules can be imported; no classes.docs.ada.cx/docs/automation/tools/code-…
Ships it: PartlyCode tools authored through the MCP Server land on a change set that only goes live when a person promotes it, so Ada stages rather than ships autonomously.Through Ada’s MCP Server, where you describe the change in plain language and it lands on a change set that goes live only when a person promotes it.ada.cx/platform/ada-computer
Done-for-you service: PartlyYou licence and run the platform yourself, but Ada also sells consultants who embed with your team to build and scale the agent, so the service half is an add-on to the software rather than the product itself.Ada's strategic and technical experts embed with your team to establish and scale your ACX function and continuously improve your AI customer service…ada.cx/experts

How much it does unattended

Runs autonomously: YesThe vendor states its agents resolve the large majority of inquiries end to end with no human in the loop, and the documented loop is decide, act, check, decide, with handoff to a human only as the fallback.It resolves 84% of customer inquiries autonomously across voice, messaging, and email channels.ada.cx/platform
Multi-agent: PartlyAda orchestrates several LLMs behind one unified Reasoning Engine rather than describing multiple cooperating agents on one task.Orchestrate multiple LLMs to understand, act, and resolve complex inquiries with accuracy, safety, and consistencyada.cx
Agent permissions: YesEvery tool is off until an administrator enables it, network access from code tools is deny-by-default against an allowed-domains list, and playbooks can require a customer confirmation step before a write runs.Connect an MCP server by URL and every tool it exposes is discovered. Each stays off until you turn it on.ada.cx/platform

Which models it runs on

Claude: PartlyClaude is supported as a client of Ada's MCP server, so you can manage and analyse the agent from Claude Desktop or Claude Code; the vendor does not disclose which LLM powers the agent itself.Most of the operations available in the Ada dashboard can now be done conversationally from Claude, ChatGPT, or any MCP-compatible client.docs.ada.cx/mcp/introduction/overview
GPT: PartlyChatGPT and OpenAI's Codex CLI are documented MCP clients for driving Ada, but that is the management surface; the models behind the customer-facing agent are never named.Codex CLI connects to the MCP server over HTTP. It does not use mcp-remote. Both OAuth and API key authentication are supported.docs.ada.cx/mcp/introduction/getting-st…
Open models: PartlyAda internally hosts open-source LLMs for contextual understanding only; reply generation runs on the commercial providers and the customer selects neither.internally hosts other open source LLMs for additional contextual understandingsecurity.ada.cx
Model choice: NoAda curates and selects the models itself, benchmarking and phasing in versions; no page offers the buyer a model choice.New LLMs are benchmarked across accuracy, safety, latency, and cost before selection. Updated versions pass a rigorous evaluation pipeline:…ada.cx/platform/trust

Where you use it

In your editor: PartlyThere is no IDE extension for end users, but the MCP server is documented as working inside Cursor and VS Code, which makes the editor a management surface for the agent rather than a place the product runs.Cursor — configure through Cursor's MCP settings. Both OAuth and API key work. VS Code — configure through MCP extensions that support remote MCP…docs.ada.cx/mcp/introduction/getting-st…
On the command line: PartlyTerminal use is limited to managing Ada through MCP from Codex CLI or Gemini CLI; Ada ships no CLI of its own, so this is a client-side surface the vendor supports rather than a product binary.Codex CLI installed and signed in. The codex mcp add command with the --url option. Run codex mcp add --help to confirm.docs.ada.cx/mcp/introduction/getting-st…
In your pipeline: PartlyThere is no pull-request surface, but code tools reviewed in a repository or CI pipeline can be deployed through the Platform API.Or through the Platform API, for code that’s already reviewed in a repository or CI pipeline.ada.cx/platform/ada-computer
In a browser: YesThe product is used through a hosted Ada dashboard on your own instance subdomain, and the agent itself is deployed as a web chat and messaging widget alongside voice and email.Double-click the endpoint URL to replace example with your Ada instance's subdomain. You can find your subdomain in the URL of your Ada dashboard.docs.ada.cx/reference/introduction/over…

Whose machine it runs on

Self-hosted: NoEvery documented deployment is a vendor-hosted Ada instance reached at its own subdomain, with Ada holding the conversation data under its own retention schedules; no self-hosted or bring-your-own-infrastructure option appears anywhere in the documentation.Ada retains the resources you make available to the AI Agent and end user data only as long as necessary to provide services. End user data is…docs.ada.cx/docs/other/data-retention
Open source: PartlyOnly the client-side chat UI library is distributed publicly on npm as @ada-cx/lovelace; the platform and Reasoning Engine are hosted with no published source.Lovelace is the design system that Ada's own messaging app is built from. It is published to npm as @ada-cx/lovelace: a library of React components,…docs.ada.cx/lovelace/overview

What it costs to run

How it meters: YesBilling is metered per conversation the agent has with an end user, not per seat or per token, with a resolution-based model offered as an alternative to enterprises that ask for it; no rate per conversation is published.With our conversation-based pricing model, you pay for every conversation your AI agent has with end users, making costs easy to predict, simple to…ada.cx/platform
Free tier: NoThere is no sign-up, free plan or self-serve trial: the only published route in is a sales-led expert review and custom demo, and the pricing path redirects to that booking form.Meet with our team of experts to explore how agentic customer experience can lift CSAT, scale support, and drive new revenue.ada.cx/demo
API access: YesA documented REST API family covers knowledge, end users, conversations, tools, playbooks, data export and configuration snapshots, with rotatable keys, versioning, a downloadable OpenAPI spec and a live playground.Ada's APIs adhere to REST standards and utilize JSON formatting for data exchange. Secure access is managed through Authentication using rotatable…docs.ada.cx/reference/introduction/over…
MCP server: YesAda runs its own MCP server over Streamable HTTP with OAuth or API-key auth, exposing most dashboard operations to any MCP client; it also consumes MCP servers you run as agent tools.Ada's MCP Server connects your AI assistant directly to your Ada instance so it can analyze performance, update and create resources (knowledge…docs.ada.cx/mcp/introduction/overview
Bring your own key: NoInference runs on Ada's own provider agreements; no bring-your-own-key or own-cloud path exists. You do supply credentials for your own systems when connecting API and MCP tools.All large language model (LLM) providers operate under Zero Data Retention (ZDR) agreements, ensuring customer data is never stored or reused by…ada.cx/platform/trust

Buying it for a team

A company can buy it: YesThe product is sold only to organisations: enterprise contracts negotiated through a sales motion, per-instance team management, and dashboard invitations that reject consumer email domains outright.You must enter a business email address. Domains like gmail.com or outlook.com are not accepted.docs.ada.cx/docs/other/team-access/my-t…
Seat model: PartlyDashboard access is per named user with four roles, invited and removed per Ada instance, but no seat price, seat minimum or seat ceiling is published anywhere, so the commercial half of the seat model is unknown.The available roles are Owner, Admin, Agent, and Read Onlydocs.ada.cx/docs/other/team-access/my-t…
Pooled budget: PartlyConsumption is metered per end-user conversation at the agent level rather than against per-user allowances, but no balance, pool or overage mechanics are published.you pay for every conversation your AI agent has with end usersada.cx/platform
Admin controls: YesAdmins constrain the agent on several axes: role-based dashboard permissions, enabling or disabling each API, code and MCP tool individually or in bulk, redaction settings, session timeouts, and availability rules that limit which resources an agent may use.Each tool shows its type (API, Code, or MCP), its availability, where it's used, and whether it's active. Search or filter by type and status to find…docs.ada.cx/docs/automation/tools
Audit log: YesA dedicated audit log records who changed what, when and through which interface across playbooks, tools, API keys, team access and SSO config, filterable and exportable to CSV up to 10,000 rows per export.The Audit log records configuration changes made to your AI Agent, giving administrators a chronological, searchable history of who changed what,…docs.ada.cx/docs/other/audit-log
Single sign-on: YesAny SAML 2.0 identity provider can be connected through Ada's Okta service-provider instance, with email, first and last name attribute mapping; OIDC is not offered, IdP-initiated login is unsupported, and setup requires an Ada representative.SAML-based authentication: Connect any SAML 2.0 compliant identity provider to Ada.docs.ada.cx/docs/other/team-access/sso

What happens to your code

Opt out of training: YesZero data retention is enforced with every LLM provider as the default rather than a setting you opt into, and the vendor states customer data is never stored by or used to train third-party models.Ada enforces zero data retention (ZDR) with all LLM providers, meaning customer data is never stored or used to train third-party AI models.ada.cx/platform/trust
Data residency: PartlyResidency options exist and agents can live in different regions, but no jurisdiction or region list is published: the terms are set in the enterprise agreement and negotiated with Ada's legal team case by case.Ada's enterprise agreement specifies data residency options, retention schedules, and deletion rights.ada.cx/platform/trust
Getting out: YesConversation and message records export through the Data Export API into your own warehouse, whole agent configurations export as a portable snapshot archive, and the audit log downloads as CSV; no credit expiry applies since no credit model is published.The Data Export API provides access to conversation and message records. You can use this API to export data into your own data warehouse for…docs.ada.cx/reference/introduction/over…
Certifications: YesThe vendor claims SOC 2 Type II and HIPAA certification, GDPR and PCI DSS compliance, and AIUC-1 certification, backed by annual third-party penetration tests including the LLM layer and a public trust centre.HIPAA, SOC 2, GDPR, PCI DSS, and AIUC-1 compliance safeguard sensitive data with industry-leading security practices.ada.cx/platform

35 sourced claims on this page. Checked 2026-10-07. How we check.

Alternatives

Where we would send a reader instead, in that listing's own words.

Decagon
Enterprise platform for building and running AI customer-support agents across chat, voice and email.
Sierra
Enterprise platform for customer-service AI agents across voice, chat, email and messaging, priced by outcome.
Cresta
Enterprise contact-centre platform with autonomous voice and digital AI agents plus real-time human agent assist.
Crescendo
AI-native CX platform: agents resolving support across voice, chat, email and SMS, with a staffed team behind them.
Fin
Intercom's customer agent that resolves support conversations across chat, email, voice and social channels.
Zendesk AI agents
Customer-service AI agents for messaging, email and voice, billed per automated resolution.

Filed under

APIAutonomousClaudeEnterpriseGPT

Sources and updates

The pages we read42

Every fact below comes from the vendor. Nothing here is independently corroborated yet.

Coverage elsewhere4

Updates6

2026-10-07Capabilities · Commercial terms · Free tier · Price
2026-10-06Capabilities · Commercial terms · Price
2026-10-04Capabilities · Commercial terms · Free tier · Price
2026-10-03Capabilities · Commercial terms · Free tier · Price
2026-09-09Capabilities · Commercial terms · Free tier · Price
2026-09-02Capabilities · Commercial terms · Free tier