> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.ibee.co.in/docs/infrastructure/cloud-vms/instance-types/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.ibee.co.in/_mcp/server.

# Instance types

> Choose the right Cloud VM instance type based on CPU, memory, disk, and workload profile.

IBEE Solutions VM instance types come in two families: **Standard** for balanced everyday workloads and **Performance** for compute-optimized workloads that need higher CPU density. Every instance type ships with included disk and bandwidth.

## Standard VMs — Network Bandwidth

Reliable virtual machines with balanced vCPU and memory — the right default for most production workloads. Each plan ships with a private-network ceiling that scales with size and a default 100 Mbps public-network NIC.

| Plan             | Private Network | Public Network     | Upgradable Public   |
| ---------------- | --------------- | ------------------ | ------------------- |
| `std.1.c2.m8`    | Up to 5 Gbps    | 100 Mbps (default) | On request to sales |
| `std.1.c4.m16`   | Up to 10 Gbps   | 100 Mbps (default) | On request to sales |
| `std.1.c8.m32`   | Up to 20 Gbps   | 100 Mbps (default) | On request to sales |
| `std.1.c16.m64`  | Up to 25 Gbps   | 100 Mbps (default) | On request to sales |
| `std.1.c32.m128` | Up to 40 Gbps   | 100 Mbps (default) | On request to sales |
| `std.1.c64.m256` | Up to 40 Gbps   | 100 Mbps (default) | On request to sales |

`std.1.c4.m16` is the most popular choice for new production deployments.

## Performance VMs — Network Bandwidth

Compute-optimized virtual machines with stronger CPU density, for workloads that need faster processing per core. Each plan ships with a private-network ceiling proportional to size and a default 100 Mbps public-network NIC.

| Plan              | Private Network | Public Network     | Upgradable Public   |
| ----------------- | --------------- | ------------------ | ------------------- |
| `perf.1.c1.m2`    | Up to 2 Gbps    | 100 Mbps (default) | On request to sales |
| `perf.1.c2.m4`    | Up to 5 Gbps    | 100 Mbps (default) | On request to sales |
| `perf.1.c4.m8`    | Up to 10 Gbps   | 100 Mbps (default) | On request to sales |
| `perf.1.c8.m16`   | Up to 20 Gbps   | 100 Mbps (default) | On request to sales |
| `perf.1.c16.m32`  | Up to 25 Gbps   | 100 Mbps (default) | On request to sales |
| `perf.1.c32.m64`  | Up to 40 Gbps   | 100 Mbps (default) | On request to sales |
| `perf.1.c64.m128` | Up to 40 Gbps   | 100 Mbps (default) | On request to sales |

`perf.1.c4.m8` is the most popular Performance plan.

### Reading the columns

| Column                | Meaning                                                                                    |
| --------------------- | ------------------------------------------------------------------------------------------ |
| **Private Network**   | Maximum bandwidth between VMs in the same region over the IBEE backplane.                  |
| **Public Network**    | Default outbound bandwidth on the public NIC. Inbound is unmetered.                        |
| **Upgradable Public** | The public NIC can be raised above 100 Mbps — contact IBEE sales to request a higher tier. |

vCPU, RAM, and disk for each plan are encoded in the plan name (see [Naming convention](#naming-convention)).

## Choosing a family

| Pick Standard if…                                      | Pick Performance if…                         |
| ------------------------------------------------------ | -------------------------------------------- |
| Workload is memory-balanced (typical web/API/database) | Workload is CPU-bound and scales with cores  |
| You want the broadest fit for production               | You need higher per-core throughput          |
| You expect mixed CPU and RAM usage                     | RAM is comfortable but CPU is the bottleneck |

For GPU-accelerated workloads (training, inference, rendering), see [GPU VMs](/docs/infrastructure/gpu-vms).

## Naming convention

Plan names follow `<family>.<gen>.c<vCPU>.m<RAM-GiB>`:

* `std.1.c4.m16` — Standard family, generation 1, 4 vCPU, 16 GiB RAM.
* `perf.1.c8.m16` — Performance family, generation 1, 8 vCPU, 16 GiB RAM.

The generation number changes when the underlying hardware platform is refreshed.

## What's included

Every plan includes:

* **NVMe root disk** sized per plan
* **Outbound bandwidth** allowance per plan; inbound is unmetered
* **Public IPv4 address** when deployed with **Public IP** networking ([Networking for VMs](/docs/infrastructure/cloud-vms/networking-for-vms))
* **Snapshot and backup eligibility** ([Backups](/docs/tools/backups), [Snapshots](/docs/tools/snapshots))
* **Hourly billing**, plus **Monthly** or **Yearly** commitment terms where the plan offers them

Storage beyond the included disk is added separately as [Block Storage](/docs/infrastructure/block-storage). Prices for each plan are shown on the Deploy page.

## Resizing between plans

You can move a VM to a different plan from **Settings → Change Plan**. The VM is briefly stopped, resources are changed, and it restarts on the new plan. Upgrades grow CPU, RAM, and disk; downgrades can reduce CPU and RAM, but the disk never shrinks. See [Resize a VM](/docs/infrastructure/cloud-vms/resize-a-vm).

## Related pages

* [Launch your first VM](/docs/infrastructure/cloud-vms/launch-your-first-vm)
* [Create a VM](/docs/infrastructure/cloud-vms/create-a-vm)
* [Resize a VM](/docs/infrastructure/cloud-vms/resize-a-vm)
* [GPU VMs](/docs/infrastructure/gpu-vms)
* [Regions and locations](/docs/getting-started/overview/regions-and-locations)

> Choose the right Cloud VM instance type based on CPU, memory, disk, and workload profile.