On June 3, 2026, Anthropic introduced the Services Track and the Claude Partner Hub, the most significant expansion of the Claude Partner Network since the program launched in March. The Claude Partner Network now formalizes a three-tier ladder for software agencies and consultancies, Select, Preferred, and Global Premier, with very specific gating requirements based on certified practitioners, joint customer deployments running in production, and public customer stories. If you run an agency, this is the clearest signal yet of how Anthropic wants its services ecosystem to look in 2026, and where the bottleneck for promotion will be: shipping real, deployed production apps for real customers.
Quick Answer
- The Claude Partner Network Services Track is Anthropic's new three-tier program for service partners, announced June 3, 2026, with tiers named Select, Preferred, and Global Premier.
- Entry to Select requires 10 active Claude-certified individuals, 2 joint customers deployed in production over the trailing 12 months, and 1 public customer story.
- Preferred and Global Premier scale those requirements by 10x and 100x respectively, plus regional spread for the top tier.
- The Claude Partner Hub is a portal that shows your standing against published requirements, refreshed daily, and helps customers find qualified firms.
- The hidden bottleneck is not certifications, it is shipping production deployments fast enough to clear the joint-customer thresholds.
What changed on June 3, 2026
When Anthropic launched the Claude Partner Network in March 2026, it was a relatively loose program: certifications, a partner directory, and access to Anthropic Partner Academy exams. By June, more than 40,000 firms had applied and more than 10,000 consultants had earned a Claude certification, according to the announcement. That kind of volume needed a structured tier system, and the Services Track is the answer.
The Services Track is meant for partners whose business is building things with Claude for customers: software agencies, system integrators, vertical consultancies, AI-services firms. It sits next to the existing Build and Solutions tracks that focus on technology partners and ISVs. The Services Track does not replace those, it sits beside them.
There are three Services Track tiers, with promotions reviewed twice a year on January 1 and July 1. For 2026, Anthropic added a third review window on October 1, which is the only chance in this calendar year to climb a tier without waiting until next January.
Inside the three Services Track tiers
These are the published requirements at announcement time:
| Tier | Active Claude-certified individuals | Joint customer deployments in production (trailing 12 months) | Public customer stories | Other |
|---|---|---|---|---|
| Select | 10 | 2 | 1 | None disclosed |
| Preferred | 100 | 15 | 3 | None disclosed |
| Global Premier | 1,000 | 100 (across 3+ regions) | 15 | Joint business plan with named executive sponsors |
A few observations from the structure:
- Certifications scale 10x per tier. That is straightforward, an agency can plan its certification calendar and hit it. The Anthropic Partner Academy publishes exam dates, and certifications are individual and renewable.
- Joint customer deployments scale faster than certifications at the top. From Select to Preferred you need 7.5x more deployed customers (2 to 15). From Preferred to Global Premier you need another 6.6x (15 to 100), plus geographic spread across at least three regions. This is the hard part. Certifications can be scheduled, customer deployments cannot.
- Public stories grow more slowly. 1, 3, 15. Most agencies will find the public-story requirement easier to meet than the deployment requirement, as long as customers consent.
- Global Premier is genuinely top-tier. 1,000 certified individuals plus 100 production customers plus a jointly developed business plan with named executive sponsors at Anthropic. This is the tier the big global SIs will fight for. Most agencies are not playing that game.
What the Claude Partner Hub actually does
The Partner Hub is the portal that backs the Services Track. According to Anthropic, it does two things:
- For partners: shows your firm's current standing against the published requirements, refreshed daily. You can see exactly how many active certifications you have, how many deployed joint customers count toward each tier, and what is missing before the next review window.
- For customers: lets buyers searching for Claude expertise filter partner firms by track, tier, certifications, and (presumably) region. Anthropic frames this as helping customers find the firms most qualified for the scope of their project.
The portal also surfaces the partner's catalog of Claude-based services. The technologies explicitly mentioned in the announcement are Claude Code, Claude Cowork, the Model Context Protocol (MCP) connector, and Anthropic Partner Academy certifications, which gives you a strong hint that successful service offerings will be packaged around those product surfaces.
The hidden bottleneck: shipping production deployments
If you are reading this as an agency operator, the bottleneck is not the certifications. You can put a calendar in front of your team and book them through the Anthropic Partner Academy.
The bottleneck is shipping enough joint customer deployments in production within a 12-month rolling window to clear the next tier. Two for Select. Fifteen for Preferred. One hundred for Global Premier. And these have to be real production deployments, not pilots, not demos.
In a typical Claude services engagement, the work breaks down roughly into three buckets:
- Discovery and orchestration design. Where do you put Claude, which tools does it call, what guardrails, what eval coverage.
- The actual application around Claude. Auth, a database, file storage, payment flows where relevant, a custom admin UI, a hosted web app or API, a custom domain, CI/CD. The boring production wiring.
- Operate, measure, iterate. Logs, billing, model upgrades, prompt revisions.
Buckets 1 and 3 are where your Claude-certified people add the most value. Bucket 2, the actual production application around Claude, is where agencies bleed weeks. That bucket is the same regardless of which Claude model the customer uses, and it is the same as every other web app your agency has ever built. It is the work that should not be the rate limiter on your tier ladder.
This is exactly where an AI app builder pairs with Claude services well.
How software agencies climb tiers faster with Claude + Totalum
Totalum is an AI app builder that creates production-grade Next.js applications with auth, database, file storage, hosting, payments, and custom domains built in. It is a peer to Lovable, Bolt, Replit, and v0, not a deploy layer behind a coding agent. Two things make it interesting for Claude services partners.
First, it has its own MCP server. Any Claude-driven agent, Claude Code in a terminal, Claude Cowork in a browser, a custom Claude API integration in your internal tooling, can connect to Totalum and orchestrate the build. The agent is the prompter and the orchestrator. Totalum is the builder that materializes the project end to end, every screen, every database collection, every API route, every deploy.
Second, the output is owned by the customer. Source code is downloadable, the data lives in a Totalum-managed database with the customer's admin panel, and there is no SPA-only constraint that would hurt SEO for client-facing apps. That makes it acceptable to enterprise customers who would refuse a vendor lock-in stack.
In practice, the workflow looks like:
- Your Claude-certified engineer scopes the Claude integration: tools, prompts, evals, eval coverage targets.
- The same engineer points their Claude Code or Claude Cowork session at Totalum via MCP. Claude (the agent) drives Totalum to materialize the surrounding application: tenant model, admin panel, billing, custom branding, deploy to a custom domain.
- The team finishes Bucket 1 work (Claude integration logic) and ships. Bucket 3 starts.
The framing matters. This is not Claude writing code that Totalum runs. The agent drives an AI app builder that materializes the entire project, which frees your Claude-certified engineers to spend their hours on the work that actually counts toward your tier differentiation.
For broader context on how Claude fits across the agent ecosystem, see our take on the best AI coding agents of 2026 and on the Claude Code Skills + Totalum integration.
Where this changes your agency's positioning
Three concrete changes once the Services Track is live:
1. You can quote a faster delivery promise for Claude work. When Anthropic's Partner Hub becomes the place enterprise buyers shop for Claude expertise, time-to-deployment will matter. Agencies that can offer "production app around your Claude integration in two weeks" win those briefs.
2. You can price the surrounding application as a fixed-scope add-on. If your Claude services arm pairs with a defined "production wiring" service line built on Totalum, you can sell it as a fixed-scope add-on (auth + database + admin + deploy) attached to every Claude project. That stabilizes margins on the AI services work.
3. You can publish customer stories faster. The Services Track requires public customer stories at every tier. Stories require customers in production. If your delivery loop is shorter, your story pipeline is shorter. This compounds.
Practical 90-day playbook for entering Select tier
Assuming you start with zero Claude-certified individuals and zero joint deployments today, June 4, 2026:
| Week | Focus | Output |
|---|---|---|
| 1 to 2 | Choose 10 engineers to certify. Block calendar time, register for Anthropic Partner Academy exams. | 10 exam slots booked. |
| 1 to 3 | Identify 4 existing customer accounts where Claude can add concrete value. Scope 2 first deals. | Two signed SOWs for Claude integration work. |
| 2 to 4 | Stand up an internal delivery template combining a Claude Code or Claude Cowork loop with Totalum via MCP. | Internal template ready, deployable, documented. |
| 3 to 8 | Execute customer engagement 1 end to end. Ship production. | Production deployment 1 complete. |
| 6 to 12 | Run all 10 engineers through Anthropic Partner Academy exams. | 10 active certifications. |
| 8 to 12 | Execute customer engagement 2 end to end. Ship production. | Production deployment 2 complete. |
| 11 to 13 | Publish 1 customer story (case study + blog + Anthropic Partner Hub entry). | 1 public story live. |
Cleared on day 90: 10 certified, 2 deployed joint customers, 1 public story. That is Select tier, ready for the next promotion review on July 1, 2027, or October 1, 2026 if you ship faster.
What this means for SaaS founders embedding Claude
If you are a SaaS founder embedding Claude inside your product rather than running a services firm, the Services Track does not apply directly, but the Partner Hub still matters. Your customers will increasingly check the hub when evaluating which Claude-based vendors to trust. Two takeaways:
- A formal partnership story will become a tangible asset for trust. Even if you join through the Build track instead of Services, get on the hub.
- The same Claude + AI app builder pattern that helps services partners ship faster also helps you launch new SKUs faster inside your SaaS. The Totalum API and MCP path is built for that embedding case.
FAQ
What is the Claude Partner Network Services Track?
The Claude Partner Network Services Track is Anthropic's tiered program for service partners, announced June 3, 2026. It has three tiers, Select, Preferred, and Global Premier, with promotion reviews twice a year and an additional October 1 review in 2026.
How do agencies qualify for the Select tier?
According to Anthropic's announcement, the Select tier requires at least 10 active Claude-certified individuals, at least 2 joint customers deployed in production in the trailing 12 months, and at least 1 public customer story.
When are promotion reviews held?
Promotions are processed twice a year, on January 1 and July 1. For the first year of the program, Anthropic added an extra review on October 1, 2026, which gives agencies a third chance in 2026 to clear a tier without waiting for January.
What is the Claude Partner Hub?
The Claude Partner Hub is the portal Anthropic launched alongside the Services Track. Partners see their current standing against published tier requirements, refreshed daily, and customers can find qualified firms by track, tier, and certifications. It is reachable from the Claude Partner Network section of anthropic.com.
Which Claude products are explicitly mentioned in the program?
The announcement names Claude Code, Claude Cowork, the Model Context Protocol (MCP) connector, and Anthropic Partner Academy certifications as the technology surface for partner service offerings. There is no public pricing or revenue share disclosed.
Does Totalum count as a Claude Partner Network partner?
Totalum is an AI app builder that pairs with Claude services partners. It has its own MCP server so that Claude Code, Claude Cowork, or any Claude-driven agent can drive it to materialize production apps end to end. Agencies use it to compress the surrounding production wiring that otherwise slows their joint-customer deployments. See the Claude Code Skills + Totalum guide and the Claude Code MCP tutorial for concrete patterns.
Ready to build with Totalum?
If you run a software agency building on Claude and want to compress the production wiring around every engagement, we are happy to walk through the API and MCP path live. Book a 30-minute call at calendly.com/totalum/30min and we will show you how a Claude-driven agent loop can materialize a full production app inside Totalum so your certified engineers stay focused on the Claude work that climbs your tier.
For background reading on the model that ships with the program, see our Claude Opus 4.8 deep dive. For broader landscape context, our Claude Code vs Codex 2026 comparison covers where Claude fits versus the OpenAI stack.
Sources: Anthropic, "Introducing the Services Track and Partner Hub of the Claude Partner Network", June 3, 2026 (anthropic.com/news/services-track-partner-hub). Coverage and analysis: Yahoo Finance, Quartz, PYMNTS, IndexBox, Investing.com (June 3, 2026).