Key facts
| API compatibility | OpenAI-compatible /v1/chat/completions; change the base URL |
| Models | 30+ models, from free chat tiers to frontier reasoning |
| Pricing | Flat monthly self-serve plans with unlimited fair-use usage |
| Free plan | plugsky-micro and plugsky-lite; no card required |
| Trial | 14-day full-access trial for higher tiers |
| Deployment | Plugsky cloud, your VPC, on-prem and air-gapped |
| Data residency | Region-locked data planes; see /data-residency for current regions |
| Local presence | No Cape Town office or data-centre claim; residency is region-based |
TL;DR
- OpenAI-compatible API: change the base URL and keep your SDK.
- 30+ models behind one endpoint, from free chat tiers to frontier reasoning.
- Flat monthly plans with no per-token billing on self-serve.
- Region-locked data planes plus VPC, on-prem and air-gapped options.
- Free plan with plugsky-micro and plugsky-lite; 14-day full-access trial.
How it works, step by step
- Create a Plugsky account and issue an API key on the free plan — no card required.
- Point your OpenAI SDK at api.plugsky.com and map your model names.
- Pick the deployment topology that satisfies your residency requirement: Plugsky cloud region, your VPC, on-prem or air-gapped.
- Replay your existing prompts and evaluations against the same test set.
- Measure latency from your Cape Town network against candidate regions before cut-over.
- Review the DPA, the SLA and the /data-residency documentation with security and legal.
- Cut production over behind a flag and keep the one-line rollback ready.
Original data
Try it yourself
Open the data residency checker →
What enterprise AI cloud means for Cape Town
Cape Town is South Africa's legislative capital, with finance, software, tourism and port activity. Teams in the area run finance, software, tourism and logistics workloads, and AI adoption follows a similar pattern: a few builders want frontier reasoning models, while most production traffic goes to smaller, cheaper models for classification, extraction and chat. Plugsky serves that mix with one OpenAI-compatible endpoint and 30+ models, so you can prototype on plugsky-micro or plugsky-lite and move up without changing SDKs.
Because the API follows the OpenAI schema, existing code, prompts and evaluations keep working — see the docs for endpoints and model names.
How data residency works for Cape Town teams
Plugsky supports region-locked data planes: requests, logs and stored artefacts stay in the region you choose instead of being routed globally by default. That is the residency mechanism — and it does not depend on a Plugsky office or data centre in Cape Town. Validate the current region list and the contractual wording in the data-residency overview, the DPA and the SLA before you sign.
Where requirements are stricter, the same API can run in your own VPC, on-prem or air-gapped, which puts the processing boundary under your control.
Deployment options and latency trade-offs
Three paths cover most Cape Town scenarios: a managed Plugsky cloud region, bring-your-own-cloud inside your VPC, or on-prem and air-gapped for classified or critical workloads. Residency and latency trade off against each other — the closer the data plane sits to your users and systems, the shorter the network path, but a strict boundary can mean higher round-trip times than a globally distributed endpoint.
Measure before committing: run the same prompts from your Cape Town network against candidate regions, compare p50 and p95, then decide. Region selection is configuration, not a rewrite.
Migrating and controlling cost from Cape Town
Migration is one line: point the OpenAI SDK base URL at Plugsky and map model names, then replay your evaluations and cut over behind a flag so rollback stays trivial.
Cost control is flat-rate on self-serve plans — unlimited fair-use usage with no per-token billing — so Cape Town teams can forecast a monthly line item rather than token spend. Start free with plugsky-micro and plugsky-lite, and use the 14-day full-access trial to evaluate larger models. Current plans are listed on the pricing page.
Honest comparison
| Capability | Plugsky | Typical global API provider | Building in-house |
|---|---|---|---|
| API compatibility | OpenAI-compatible — change the base URL | Usually compatible, varies by model | Full rewrite and integration work |
| Models | 30+ models behind one endpoint | Mainly the provider's own catalogue | You host and maintain each model |
| Pricing | Flat monthly self-serve plans, no per-token billing | Per-token billing, harder to forecast | GPU, operations and staffing costs |
| Data residency | Region-locked data planes; VPC, on-prem, air-gapped options | Limited region choices | You own the responsibility |
| Local presence in Cape Town | No office or data-centre claim; residency is region-based | Varies; few publish local commitments | Depends on your own facilities |
| Migration effort | Base URL plus model mapping | Depends on compatibility gaps | Long integration cycle |
Frequently asked questions
Does Plugsky have a data centre or office in Cape Town?
No. Plugsky does not claim a local office or data centre in Cape Town; residency is delivered through region-locked data planes, or through deployments in your own VPC, on-prem or air-gapped. See /data-residency for the current region list.
How does data residency work for Cape Town teams?
You select a region and Plugsky keeps the data plane locked to it, so requests, logs and artefacts are not routed globally by default. Validate the region list and the contractual terms in the DPA and the SLA before you commit.
Can we keep using the OpenAI SDK?
Yes. Plugsky exposes an OpenAI-compatible chat completions endpoint, so you change the base URL and model name and keep your existing SDK, prompts and evaluations.
Is there a free plan?
Yes — the free plan includes plugsky-micro and plugsky-lite with no credit card required. A 14-day full-access trial is available when you want to evaluate larger models.
How is pricing structured?
Self-serve plans are flat monthly with unlimited fair-use usage and no per-token billing. See the live pricing page for current plans and fair-use limits.
Which capabilities are live today?
Chat, streaming, JSON mode, function calling, embeddings, RAG and agents are live. Audio, images, moderation, files, batch, fine-tuning, assistants and responses are coming soon — check the docs before planning those workloads.
What about latency from Cape Town?
It depends on the region you select and how your network routes to it. Measure candidate regions from your own environment, compare p50 and p95, and treat residency and latency as an explicit trade-off.
How much migration work is involved?
For a standard chat application the code change is a base URL plus model mapping; the real work is replaying your evaluations and doing a staged cut-over.