Models

What is plugsky-lite and when should you use it?

plugsky-lite is one of the two free models on Plugsky's free plan, in a catalogue of 30+ models, alongside plugsky-micro. It is built for fast, high-volume work: simple chat, classification, extraction, tagging and tool-calling loops. It supports streaming, JSON mode and function calling on the OpenAI-compatible API with a 128K-class context window. Start free with no card, then move up when a task needs more depth.

Key facts

Model classFree-plan model in the Plugsky catalogue
Best forFast chat, classification, extraction, tagging and tool loops
Free planOne of the two free models, together with plugsky-micro; no card required
Context class128K-class window; live limits are published per model
CapabilitiesStreaming, JSON mode and function calling
APIOpenAI-compatible /v1/chat/completions — no SDK changes
ResilienceAutomatic backup upstream plus same-profile fallback peers
Product statusLive

TL;DR

  • A free model on the free plan — no card required to start.
  • Built for speed and volume, not deep multi-step reasoning.
  • Supports streaming, JSON mode and function calling like paid models.
  • Pair with plugsky-micro to compare quality against latency for free.
  • Upgrade to paid models only when your evaluation shows you need to.

How it works, step by step

  1. Create a Plugsky account and generate an API key on the free plan — no card needed.
  2. Read the live model card at /models for the current window and feature flags.
  3. Point your OpenAI-compatible client at the Plugsky base URL.
  4. Set the model to plugsky-lite and run your real prompts.
  5. Compare against plugsky-micro on the same inputs to see the free-tier trade-off.
  6. Adopt a paid model for the tasks that fail your quality bar, keeping lite as the default where it passes.
  7. Monitor usage and fallbacks through your dashboard and the status page.
1Create a Plugskyaccount andgenerate an API key2Read the live modelcard at /models forthe current window3Point yourOpenAI-compatibleclient at the4Set the model toplugsky-lite andrun your real5Compare againstplugsky-micro onthe same inputs to6Adopt a paid modelfor the tasks thatfail your quality

Try it yourself

Open the LLM cost calculator →

What plugsky-lite is

plugsky-lite is one of the two models available on Plugsky's free plan. It is designed for fast, straightforward requests: short chat answers, classification, field extraction, tagging and tool-calling steps inside simple agents. Like every model in the catalogue it speaks the OpenAI-compatible chat API, with streaming, JSON mode and function calling available.

Free-tier routing is updated over time, so check the live card at /models for the current engine, context window and capability flags. That page also shows its sibling plugsky-micro, the second free model.

When to choose it

Choose plugsky-lite when you are prototyping, running low-stakes high-volume jobs, or serving simple features where a fast answer is enough. It is a genuinely useful production model for the easy majority of requests, and a zero-cost way to validate an API integration before committing to a plan.

Move up when a task needs deeper reasoning, longer synthesis or higher accuracy on ambiguous inputs. A practical pattern is to keep lite as the first hop and route failures or low-confidence results to a paid model such as plugsky-pro or plugsky-frontier. Use the 14-day full-access trial to test those paid models on your workload.

Free models: what to expect

Free tiers are for speed and access, not for the hardest tasks. Set expectations accordingly and design your system to fail over to a paid model when quality matters.

  • Good fits: FAQ chat, classification, tagging, short summaries, developer smoke tests.
  • Poor fits: long-form analysis, complex maths, multi-document synthesis, high-stakes output.
  • Use JSON mode plus validation for anything structured.
  • Keep prompts short and explicit; small models reward clarity.
  • Compare lite and micro on your own prompts — the right free model depends on your workload.

How to switch to plugsky-lite

Switching is a base URL plus a model name: {"model": "plugsky-lite", "messages": [{"role": "user", "content": "Classify this message."}], "stream": true} posted to https://api.plugsky.com/v1/chat/completions.

Keep your SDK and prompts, run your tests, and you are live. When your evaluation shows a task outgrowing the free tier, change one model name to a paid model — the integration does not change.

Honest comparison

Dimensionplugsky-liteplugsky-microPaid workhorse model
PlanFree, no card requiredFree, no card requiredPaid plan
Best fitFast simple chat and high-volume tasksFree tasks needing more reasoningGeneral production workloads
Latency profileTuned for speedStill fast, less speed-focusedBalanced
Tools and JSON modeSupportedSupportedSupported
Context class128K-class128K-class128K-class
FailoverAutomatic peer fallbackAutomatic peer fallbackAutomatic peer fallback

Frequently asked questions

Is plugsky-lite really free?

Yes. The free plan includes two free AI models, plugsky-micro and plugsky-lite, and it does not require a credit card to start.

What is the difference between plugsky-lite and plugsky-micro?

Both are free. Lite is positioned for fast, high-volume simple work; micro is the free option when a task benefits from more reasoning. Test both on your prompts.

What context window does it have?

It is 128K-class today, but the exact live window and output limit are published per model at /models. Check there before relying on a number.

Can I use it for production?

Yes, for tasks it passes on your evaluation. Keep an escalation path to a paid model for requests that need deeper reasoning or higher accuracy.

Does it support function calling and JSON mode?

Yes — streaming, function calling and JSON mode are part of the shared OpenAI-compatible surface.

Is there a trial for paid models?

Yes. A 14-day full-access trial lets you evaluate paid models on your workload before deciding.

How do I migrate existing OpenAI SDK code?

Change the base URL to the Plugsky endpoint and set the model name. Prompts, tools and response parsing stay the same.

What happens if the free model has an incident?

Requests retry through a backup upstream and same-profile fallback peers. Live health is on the status page.