Indonesia’s Indosat Ooredoo Hutchison and application-security company F5 have launched the Indosat Intelligence Suite, an enterprise platform designed to provide governed access to local and global artificial intelligence models through a single interface.
The platform became available on October 9 and combines model access, token metering, workload routing, security controls and data-residency policies. Indosat said it is initially targeting 50 businesses and organizations, although it did not identify customers or disclose commercial commitments.
One interface for local and global models
The official Intelligence Suite site lists more than 100 models in its live catalog and describes more than 300 open and closed models as part of the roadmap. The direct press announcement described access to more than 300 models, so availability may vary by tier and rollout stage.
Users can connect to models through a governed application programming interface rather than managing separate integrations for every provider. The platform is designed to authenticate requests, apply rate limits, meter token use and consolidate usage data and spending controls.
Available models include Indosat’s Sahabat-AI family, other models hosted in Indonesia and connected global services. The product site lists examples from Anthropic, Google, DeepSeek and Meta, while emphasizing that the catalog will grow over time. Model availability, pricing and routing rules may differ based on customer requirements and data-location policies.
Indosat is offering pay-as-you-go access, unified enterprise credits and dedicated capacity. The company said credits can be allocated across models, applications and teams, giving enterprises a consolidated way to track consumption. Promotional early-access pricing is planned through the year, but full commercial terms were not disclosed in the announcement.
F5 provides workload and security controls
F5 is supplying infrastructure software for tenant isolation, container security and GPU-aware load balancing. The companies said those capabilities are intended to route workloads according to factors including cost, model quality, latency and available compute capacity.
The platform also incorporates F5 AI Security capabilities intended to address model-facing threats, data leakage and application programming interface risks. These are vendor-described controls; the announcement did not provide independent performance testing, customer deployments or audited security results.
For organizations with data-residency requirements, Indosat said Sahabat-AI and its locally hosted tiers can process workloads in Indonesia, while policies can determine which requests may be sent to global models. Requests can be authenticated, logged and metered to support governance and audit requirements.
Launch builds on Indosat’s AI infrastructure push
The Intelligence Suite extends Indosat’s broader AI strategy, which spans connectivity, data centers, GPU infrastructure and locally developed models. In August, Indosat, Ooredoo Group, Nokia and NVIDIA launched the Zankore AI infrastructure platform, targeting 1 gigawatt of NVIDIA-based AI factory capacity over time.
The new suite is intended to sit above that infrastructure and give enterprises a common way to consume models and computing resources. Its token marketplace is listed as live, while the managed AI Factory Platform is scheduled for November 2026, indicating that some elements remain in phased rollout.
Indosat and F5 have not disclosed the investment value, revenue-sharing arrangement or length of their agreement. They also did not name the first 50 target organizations or provide adoption commitments. The significance of the launch will depend on customer uptake, the number of models that become generally available and whether the platform can meet its stated security, availability and data-residency goals in production environments.