In 2026, the scheduled retirement of Open CTI transforms Salesforce telephony into a major strategic issue, far beyond a simple softphone question. Between Service Cloud Voice, Aircall, Ringover, and the “Salesforce-native” players, the choices organizations make this year will shape their costs, architecture, and agility until after 2028.
Open CTI: the official end of an era
Salesforce has officially placed Open CTI inmaintenance modeand announced its withdrawal for theFebruary 28, 2028. The framework is no longer available for new Agentforce Service orgs, although existing orgs can still use it during the transition period.
Open CTI is a JavaScript API that allows CTI vendors to integrate a softphone into Salesforce Call Center, with click-to-call, screen-pop, and call control directly in the console, without a native adapter installed on the workstations. Historically, most of the “Salesforce CTI” telephony connectors were built on this component, which is now considered legacy.
Why is Salesforce retiring Open CTI now
For Salesforce, the stated goal is to provide aunified, secure, and AI-driven agent experienceby directing customers to Service Cloud Voice. Voice is treated there as a full-fledged Omni-Channel, with a native VoiceCall object, unified routing, real-time transcription, and recommendations directly in the agents' workflow.
In the background, this switch transforms a free framework into amonetized product: Service Cloud Voice becomes a high-value add-on, with specific licenses and sometimes telecom consumption via Amazon Connect or voice partners. The centralization of voice data in Salesforce also fuels the AI strategy (Agentforce, conversation analysis), which is difficult to achieve when most recordings and metrics remain stuck in external CTI portals.
2026: the pivotal year for Salesforce CTI
Old CTI approaches (CTI Toolkit, Classic integrations) have already been deprecated, and Open CTI was until now the recommended modern standard. Now, even this standard has an end-of-life date, forcing admins and architects to rethink their Salesforce telephony as early as 2026 rather than waiting until 2027–2028.
Market experts remind us that delaying migration increases the risks of unpatched vulnerabilities, incompatibilities, and ultimately service disruptions when a framework actually reaches End-of-Life. Organizations that rebuild their CTI architecture now can take advantage of this to gain flexibility, reduce dependencies, and align their telephony with their long-term data and AI vision.
Service Cloud Voice: the vision 'voice = native CRM channel'.
Service Cloud Voice positions voice as aService Cloud channelrather than a telephony silo next to the CRM. Calls become VoiceCall recordings, routed via Omni-Channel, enriched by Agentforce (transcription, sentiment analysis, summaries, Next Best Action) and visible in the same reports as other customer interactions.
On the pricing side, Salesforce markets Voice as an add-on: for example, through Voice plans with Amazon Connect including 750, 2,000, or 5,000 minutes per agent per month, or through Voice licenses for telecom partners, around €50–200/user/month depending on the included volume, in addition to Service/Agentforce licenses. This positioning makes it a particularly relevant choice for thestructured contact centers, where data consistency and AI justify the additional investment.
Aircall and Ringover: kings of sales-driven tested by 2028
Aircall: ease of use and competitive cost
Aircall has established itself as a popular cloud telephony solution for sales teams, with a Salesforce integration allowing click-to-call, call pop-ups, and automatic activity logging in Lightning. Its plans are listed around $30–50/user/month (annual) for the Essentials and Professional offerings, with AI and analytics add-ons billed a la carte.
For asales-drivenclient without strong AI requirements, these rates often remain more attractive than the combination of Service/Agentforce licenses + Service Cloud Voice, for equivalent functional use. The point of caution: part of the current Salesforce integration relies on an integrated softphone that, historically, depended on Open CTI in the utility bar, which requires a technical evolution before 2028.
Ringover: multi-channel telephony, SMS, WhatsApp
Ringover offers cloud telephony aimed at SMEs with Salesforce integration covering calls, SMS, and WhatsApp, contact synchronization, and automatic activity creation. Its plans start around $21–22/user/month for the entry-level and go up to about $44–54/user/month for more comprehensive offers, excluding additional omnichannel/AI modules.
The Salesforce integration documentation clearly shows the use of an 'Open CTI Software' in the Lightning utility bar to display the Ringover softphone, which implies a direct dependency on Open CTI for the UI part in many existing deployments. As with Aircall, the key question is not 'does it work today?', but 'what will the Ringover architecture compatible with Salesforce look like in 2028 and beyond?'.
The 'Salesforce-natives': Natterbox, PhoneIQ, and the royal road to Voice
Some players have anticipated the movement by positioning themselves asnatively Salesforceand/or strategic partners of Service Cloud Voice. Natterbox, for example, highlights a 100% Salesforce-integrated telephony, with a clearly defined migration path to 'Natterbox-powered Salesforce Voice' experiences when customers want to switch to the SCV architecture.
PhoneIQ emphasizes a CTI architecture built specifically for Lightning, Omni-Channel, and Agentforce, combining the flexibility of third-party CTI and the depth of Salesforce integration (real-time data, analytics, native reporting). These players, because they already align with VoiceCall, unified routing, and AI Service Cloud, are better equipped for the post-Open-CTI era than pure 'overlay softphones'.
2028: the real risk is not the outage, but the lock-in
The Salesforce announcement mentions the withdrawal of Open CTI on February 28, 2028, with the end of support and the risk of malfunctions if integrations are not migrated in time. Partners like Natterbox confirm that they will maintain Open CTI support until that date, but will then need to transition their clients to other architectures (native alternative or SCV-powered).
The biggest danger for clients is not a sudden "switch off," but thetrap of dependency: staying with a vendor whose only Salesforce history is Open CTI, without a concrete plan for VoiceCall, Service Cloud Voice, or independent Lightning components, condemns one to a forced migration project, potentially costly and rushed by 2027–2028.
Sales-driven clients: why third-party CTI often wins on pricing
For a sales-oriented client, with little or no AI requirements on voice, the economic comparison is compelling in many cases. A Salesforce salesperson already needs a Sales or Service license, and adding Service Cloud Voice often represents an additional 50–200 €/month per agent, depending on the tiers of minutes and the chosen telecom configuration.
In contrast, Aircall or Ringover offer plans between $21 and $50/month per user, including telephony and a Salesforce integration that is more than sufficient for use cases of prospecting, opportunity tracking, and customer follow-ups, without imposing a complete contact center architecture project. In this context, investing in Service Cloud Voice solely to "make outbound calls from Salesforce" often resembles over-equipping that is poorly aligned with the actual need.
AI-driven contact centers: where Service Cloud Voice becomes evident
As soon as we talk aboutmultichannel contact center, with serious requirements for SLA, quality, auditing, and cross-channel reporting, the balance quickly tips in favor of Service Cloud Voice. SCV enables unified routing (Unified Routing, Omni-Channel Flows) that takes into account skills, case context, and advanced business rules to direct calls to the right agent or queue.
The product developments for 2025–2026 around Service Cloud Voice, coupled with Agentforce (automated summaries, coaching, sentiment analysis, autonomous agents), enhance this value for organizations that envision their customer service over a 5–10 year horizon. In these contexts, the additional cost per agent is justified by productivity gains, a better customer experience, and much more robust data governance.
How to choose in 2026: a simple yet strategic framework
A pragmatic way to decide is to ask four questions in a scoping workshop:
Where should routing live?
If the answer is “in Omni-Channel, with centralized rule management for all channels,” we are leaning towards Service Cloud Voice or a native CTI aligned with SCV. If routing can remain in the CTI engine (sales-driven scenarios, outbound campaigns), an Aircall or Ringover remains fully legitimate.
What is the AI & data vision?
If voice is to feed Data Cloud, Agentforce, and global customer 360 analytics, Service Cloud Voice and its native partners have a structural advantage. If AI remains confined to the CTI itself (summaries, local transcriptions) and Salesforce is “only” the CRM, a third-party CTI may suffice.
What is the time horizon?
For a 3–5 year vision, with a strong likelihood of increased customer service demand, it is better not to underestimate the future migration effort and to integrate Service Cloud Voice (or an aligned partner) into the target trajectory. For a shorter horizon or a very sales-centered context, a modern third-party CTI, already moving away from Open CTI, is often optimal.
What is the acceptable budget per agent?
If the psychological ceiling is around a few dozen euros/dollars per month for voice, SCV will be difficult to justify outside of advanced service use cases. If the organization is willing to invest 100–200 €/month per agent for an AI-powered contact center position, then SCV becomes a logical building block of the overall architecture.
How to smartly prepare for the post-Open-CTI era
The best organizations do not just “choose a CTI,” theydesign a trajectory :
They audit their dependency on Open CTI (current or potential) and challenge the vendors' roadmap on their 2028+ architecture (Lightning-first, VoiceCall, SCV-ready).
They build a business case that truly compares thetotal costs over 3–5 years: licenses, minutes, AI add-ons, migration projects, training, support.
They do not sacrifice long-term architecture for a "quick win Open CTI" that risks blowing up in flight by 2028.
Anticipating the transition: The audit as a prerequisite
The most resilient organizations do not just choose a tool; they design a trajectory. It starts with a rigorous audit of the technical debt related to telephony.
In this regard, expert structures such as CRM Dojo have developed in-depth knowledge of these transition issues. Through dedicated audit tools, they allow for a precise assessment of your dependency on Open CTI and to build a comparative business case (licenses, consumption, migration costs) over a 3 to 5 year cycle.
The goal is to avoid the technical "quick win" that would turn into a financial deadlock by 2028.
Conclusion: 2026, the year when voice becomes a CRM architecture choice
The retirement of Open CTI is not bad news; it is a turning point that forces us to make explicit choices that have long been postponed: where voice lives, who governs it, how it feeds AI, and how much it really costs.
In 2026, the "best" choice is neither Service Cloud Voice nor Aircall nor Ringover in absolute terms: the best choice is the one that alignsarchitecture, business model, and data visionon the horizon of 2028 and beyond, without locking the company into a technical or financial dead end. Organizations that treat voice as a strategic component of their Salesforce platform – and not just as a simple dialer – will be the ones that truly benefit from this historic turning point in CTI.