canonical: https://jentic.com/apis/gcore.com/gcore # Gcore OpenAPI The Gcore API is the unified control plane for Gcore's global cloud platform - covering compute (bare metal, virtual instances, GPU, containers, FaaS, managed Kubernetes), networking (load balancers, floating IPs, security groups), CDN, DNS, edge storage, video streaming, AI inference (Everywhere Inference), DDoS protection, and a WAAP (web application and API protection) suite. With 1,123 endpoints across 130+ tag groups, it gives agents end-to-end provisioning and operational control of multi-region cloud workloads. Authentication uses an API key passed in the Authorization header. ## For AI agents Provision and manage Gcore cloud compute, networking, CDN, DNS, video streaming, edge AI inference, and DDoS protection from a single API. ## Scope Does not handle accounting, CRM, or end-user authentication - use for cloud, edge, CDN, DNS, streaming, and security infrastructure operations only. ## Capabilities - Provision virtual instances, bare metal, and GPU servers across Gcore regions through the Instances and Bare Metal endpoints - Configure CDN resources, origins, rules, and SSL certificates to accelerate static and dynamic content - Manage authoritative DNS zones, RRsets, and DNSSEC keys for production traffic - Deploy and scale containerised workloads with Container as a Service and Managed Kubernetes - Run AI workloads at the edge via Everywhere Inference apps and FastEdge functions - Apply DDoS protection profiles and WAAP firewall rules to public-facing domains - Stream live broadcasts and on-demand video with the Streams, Broadcasts, and Videos resources ## Use cases ### Multi-Region Cloud Provisioning Spin up a multi-region application footprint - VMs, load balancers, security groups, and floating IPs - from one API surface. The agent reads `/cloud/v1/regions`, picks targets, and chains Instance, Network, and Load Balancer creates. Replaces multi-vendor scripts with a single Gcore control plane. Example prompt: List regions, then create an instance with the chosen region id, attach a floating IP, and register it behind a load balancer through the Instances and Load Balancers endpoints ### Edge AI Inference Deployment Deploy and route AI inference at the edge for low-latency consumer applications. The agent uses Everywhere Inference Apps endpoints to register a model, picks regions, and exposes it through Gcore's edge network. Cuts the cold-start and round-trip latency seen with centralised inference. Example prompt: Use the Everywhere Inference Apps endpoints to create an app, attach the model, and select edge regions; then route traffic by updating the relevant configuration resources ### CDN And DDoS Protection For A New Domain When launching a new domain, agents wire up the CDN resource, attach an SSL certificate, point DNS records at the edge, and enable DDoS protection - all in one workflow. Removes the manual hand-off between CDN, DNS, and security tooling. Example prompt: Create a CDN resource pointed at the origin, attach an SSL certificate, create RRsets in the DNS zone, then attach a Protection Profile under DDoS Protection ### Video Streaming For Live Events Spin up a low-latency live stream for an event, including ingest, transcoding qualities, recording, and a player. The agent uses Broadcasts and Streams endpoints to create the stream and pulls Streaming Statistics during the event for monitoring. Example prompt: Create a Stream, attach a QualitySet, create a Broadcast referencing the stream id, and poll Streaming Statistics during the event ### Agent-Driven Cloud Operations Through Jentic AI agents discover the right Gcore operation among 1,123 endpoints using Jentic's intent search, instead of crawling the developer site. Jentic isolates the API key in its vault and returns operation schemas so an agent can provision a server, push a CDN rule, or roll out an edge function from a single search-load-execute flow. Example prompt: Use Jentic search query 'provision a virtual server' to load the gcore_create_instance tool and execute with region, flavor, and image parameters ## Key resources - **Instances** — Provision and manage virtual machines - **Bare Metal** — Provision dedicated bare-metal servers - **CDN service** — Configure CDN resources, origins, and rules - **Zones** — Manage authoritative DNS zones and RRsets - **Load Balancers** — Manage load balancer pools and listeners - **DDoS Protection** — Apply protection profiles to networks and domains - **Everywhere Inference** — Deploy edge AI inference apps and models - **Streams** — Manage live and on-demand video streams ## Why Jentic - **Setup:** Wiring the Gcore OpenAPI by hand means setting up its API-key header auth against api.gcore.com and integrating across its cloud, edge, CDN, DNS, streaming, and security operations yourself. Through Jentic you install once, import the Gcore OpenAPI from the API Directory, store the key once, and your agent calls it. - **Permission scoping:** Gcore spans many services, so scope this by operation: limit the agent to the operations it needs, such as DNS or CDN reads for one service, and leave the rest of the surface out of the allowed set. Every operation the agent can run is one you chose to include. - **Credential handling:** Your Gcore API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context. - **Discovery method:** Agents search Jentic by intent such as 'create a DNS record' or 'purge a CDN cache', and Jentic returns the matching Gcore operation with its input schema so the agent calls the right endpoint without browsing the reference docs. ## Related APIs - **DigitalOcean API** — DigitalOcean offers a simpler developer-focused cloud; Gcore covers a wider footprint including edge AI, CDN, and DDoS protection in one API. - **Cloudflare API** — Cloudflare overlaps on CDN and DDoS protection; teams sometimes use it alongside Gcore for split-traffic or fallback strategies. - **Linode API** — Linode (Akamai) provides VM and Kubernetes infrastructure; Gcore's API additionally exposes CDN, streaming, and edge AI in the same surface. ## FAQ ### What authentication does the Gcore API use? Gcore uses an API token passed in the Authorization header. Tokens are issued under API Tokens in the Gcore portal. Through Jentic the token is stored encrypted in the vault and injected per request. ### Can I provision a virtual server with the Gcore API? Yes. Create one through the Instances endpoints by specifying a region id, flavor, image, network, and SSH key. Bare metal, GPU bare metal, and GPU virtual instances each have their own dedicated tag groups under the same API. ### Does the Gcore API support edge AI inference? Yes. The Everywhere Inference and Everywhere Inference Apps endpoints let you register models, create inference apps, and route traffic to Gcore's global edge so requests resolve at the closest PoP. ### What are the rate limits for the Gcore API? Gcore enforces per-token and per-account limits that vary by product (cloud, CDN, DNS, streaming). The OpenAPI does not declare them inline; back off on HTTP 429 and consult your Gcore plan tier for hard ceilings. ### How do I configure a CDN resource through the Gcore API? Create a CDN resource via the CDN resources endpoints with an origin reference, then attach SSL certificates, configure rules under CDN rules, and route traffic by adding the corresponding RRsets in the DNS zone. ### How do I provision Gcore infrastructure through Jentic? Install with pip install jentic, search 'provision a virtual server', load the schema for the gcore_create_instance operation, and execute with region, flavor, and image. Jentic injects the Authorization header from the vault. ### Can I limit what my agent is allowed to do with the Gcore API? Yes. Because Jentic One is self-hosted by you, your own rules decide which Gcore operations and credentials the agent may use, and the Gcore surface spans over 1,123 operations so you scope by operation. You can allow only the endpoints a task needs, such as read-only access to DNS Zones or CDN resources, and leave provisioning, DDoS Protection, and Instances endpoints out of the allowed set. Every operation the agent can call is one you chose to include, and your Gcore API key is injected at execution time rather than exposed to the agent.