Key facts
| Vision input | Image understanding runs on the OpenAI-compatible chat completions endpoint with supported multimodal models |
| Example model | plugsky-gemma-4 handles text plus image inputs such as screenshots, forms and product photos |
| Model access | 30+ models from efficient to frontier tiers behind one key |
| Dedicated media endpoints | Image generation and editing endpoints are labelled coming soon |
| Deployment | Hosted, VPC, on-prem and air-gapped options on Enterprise |
| Data handling | Data encrypted at rest and in transit; API data is not used to train models |
| Audit | Per-request logs with model, tokens, latency and user; SIEM export; retention up to 7 years |
| Free plan | plugsky-micro and plugsky-lite, no card; 14-day full-access trial |
TL;DR
- Ship a photo feature with the chat call you already use.
- Narrow questions with strict schemas are measurable; open prompts are not.
- Resize images before sending to keep costs down.
- Start with an efficient multimodal model and escalate only hard cases.
- Prototype free; image generation is coming soon, so plan a fallback.
How it works, step by step
- Pick one user problem that a single photo answers.
- Define a strict output schema for the fields or labels you need.
- Call the vision endpoint with a resized image and validate server-side.
- Show results with a correction or retake path for low-confidence cases.
- Measure accuracy and correction rates on real user images.
- Escalate hard cases to a stronger model before upgrading everything.
- Add scoped keys and audit logging before your first enterprise conversation.
Original data
Try it yourself
Open the best AI model selector →
Why startups need image understanding
Photo-powered features are one of the cheapest ways to make a product feel modern: snap a receipt, a form, a shelf, a plant, a damaged part, and get something useful back. The engineering trap is treating vision as a new platform. It is not — it is the same chat call with an image in the message.
Start with a narrow question and a strict output. 'Extract these four fields from this receipt' beats 'describe this image' because you can measure it, show it in a UI and decide failure cases honestly.
How vision works on the Plugsky API
The vision path is a chat completion with an image in the message content. Send the image alongside the prompt using the OpenAI-compatible request shape you already use for text, and the model returns a normal completion that you can stream, constrain with JSON mode or feed into function calling. plugsky-gemma-4 is the workhorse for high-volume visual triage — screenshots, forms, product photos and document images — while frontier models handle harder reasoning over the same input. For startups, that means a shippable feature in days: one call, one schema, one queue, and a model tier you can change with a string.
The controls startups should configure first
Keep the MVP simple: one endpoint, one schema, one queue. Resize images client-side or server-side before sending, because resolution drives token usage. Validate every field server-side and surface a retake option when confidence is low, which is cheaper than support tickets.
On cost, begin with an efficient multimodal model such as plugsky-gemma-4 for routine images and escalate only hard cases; 30+ models sit behind one key, so switching tiers is a model-name change. Prototype on the free plan, and remember image generation endpoints are labelled coming soon — if your roadmap needs generation, keep a fallback.
A practical pilot path
Build the smallest thing that demonstrates value: upload, extract, display, correct. Measure accuracy on your own images and correction rates in the UI, then decide whether to invest further. See the live pricing page for current plans.
Honest comparison
| Capability | Plugsky | Vision-only API vendor | Building in-house |
|---|---|---|---|
| Image understanding | Live on supported multimodal chat models | Usually the core product | You host and serve vision models |
| Image generation | Labelled coming soon | Often available today | Custom deployment work |
| API shape | OpenAI-compatible chat completions | Vendor-specific SDK | You standardise it yourself |
| Model choice | 30+ models behind one key | Narrower catalogue | One model per integration |
| Deployment | Hosted, VPC, on-prem and air-gapped | Usually hosted only | Your infrastructure only |
| Controls | Region pinning, BYOK, per-request audit logs, SIEM export | Varies by vendor | You build the pipeline |
Frequently asked questions
Can we build a vision MVP on the free plan?
Yes. The free plan includes plugsky-micro and plugsky-lite with 2 API keys and no card, which is enough to validate an image feature with real users. A 14-day full-access trial supports wider testing.
Which model should we start with?
Start with an efficient multimodal model such as plugsky-gemma-4 for routine images, measure accuracy on your own set, and escalate only the cases it fails.
Do we have to change our application code?
No. Image understanding uses the same OpenAI-compatible chat completions endpoint, so your SDK, auth and error handling carry over; the message content simply includes an image.
Is there a free plan?
Yes — plugsky-micro and plugsky-lite with 2 API keys and no credit card. A 14-day full-access trial is also available.
How is pricing structured?
Self-serve plans are flat monthly with unlimited fair-use usage and no per-token charges or overage fees. See the live pricing page for current plans.
Is image generation available?
Not yet — image generation and editing endpoints are labelled coming soon. Image understanding is live today on supported multimodal chat models.