Key facts
| API compatibility | OpenAI-compatible /v1/chat/completions (change the base URL) |
| Models | 30+ models from free to frontier tiers behind one API |
| Agent primitives | Function calling, JSON mode and streaming are live |
| Retrieval | Embeddings and RAG over your own corpus |
| Deployment | Plugsky cloud, VPC, on-prem or air-gapped |
| Pricing | Flat monthly self-serve plans with fair-use usage; see the live pricing page |
| Document handling | Pass document text as context; the files endpoint is coming soon |
| Audit | Per-request logs plus tool-call and approval records |
TL;DR
- Keep your OpenAI SDK — change the base URL and model name.
- 30+ models behind one API, from free chat models to frontier reasoning.
- Deployment options from hosted cloud to VPC, on-prem and air-gapped.
- Keep determinations and payments with licensed staff.
- Evidence every agent action with logs before scaling.
How it works, step by step
- Define the job, the permitted data sources and where a human must approve.
- Map claim types and define which steps need human review.
- Create a Plugsky account and generate an API key (free plan, no card required).
- Point your OpenAI SDK at the Plugsky base URL and map your model names.
- Index the approved corpus with embeddings and keep retrieval role-scoped.
- Measure classification accuracy on a sample of real claims before go-live.
- Measure quality on your own samples, then scale with usage monitoring.
Try it yourself
Open the AI data residency checklist →
Where AI agents pay off in insurance
Insurance teams do not lack ideas for agents; they lack a safe path from demo to production. The pattern below targets repetitive, document-heavy work where a human can check the output, which is where agents earn their place first. Treat the agent as a new team member with a narrow brief, explicit permissions and a probation period, and rollout becomes an operations exercise rather than a leap of faith.
- Claims intake — classify new claims and request missing information
- Policy Q&A — answer coverage questions with citations to wording
- Underwriting research — summarise submissions and flag missing data
- Customer updates — draft status messages for adjuster approval
A reference architecture for insurance agents
An intake agent checks a claim against your checklist through tools and drafts the missing-information request, while a research agent retrieves clause text and cites it. Decisions, payouts and coverage determinations remain with licensed staff.
- Gateway with per-team keys and strong audit logging
- Retrieval over policy wording, procedures and FAQs
- Tools into claims, CRM and document stores
- Approval gates before any external communication
Data governance and human oversight
Insurance work is evidence-heavy and regulated. Keep customer and claims data inside the boundary, log every agent step, and ensure outputs that affect a customer are reviewed by a qualified person. Validate specific obligations with your compliance team.
- Role-based retrieval and tool scopes
- Immutable audit trails for each agent action
- Human review for determinations and payments
- Retention aligned with your records policy
From pilot to production
Start with internal research and drafting where a reviewer catches errors. Move to intake triage once extraction and classification accuracy is measured on your own claim mix.
Keep the rollout reversible: run the agent in shadow mode alongside the current process, compare outputs on your own samples, and move it into the workflow only when the evidence holds. Document what you measured so expanding to the next team is a decision, not a hope.
Honest comparison
| Capability | Plugsky | Typical cloud AI API | Building in-house |
|---|---|---|---|
| API compatibility | Drop-in base URL change | Usually compatible | Full rewrite |
| Model access | 30+ models behind one API | Vendor's own catalogue | You host each model |
| Pricing | Flat monthly self-serve plans; see live pricing | Often per-token | GPU + ops cost |
| Deployment | Cloud, VPC, on-prem or air-gapped | Usually vendor cloud regions | You own the stack |
| Regulated decisions | Human-owned; agents assist | Varies | You enforce it |
| Audit trail | Per-request logs plus approval records | Varies | You build it |
Frequently asked questions
Do we have to rewrite our application?
No. The chat completions API is OpenAI-compatible, so you change the base URL and model name and keep your existing SDK.
Is there a free plan?
Yes — the free plan includes two free AI models, plugsky-micro and plugsky-lite, with no credit card required.
How is pricing structured?
Self-serve plans are flat monthly with fair-use usage and no per-token charges; see the live pricing page for current plans.
Which endpoints 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 endpoints are coming soon — check the docs before planning around them.
Can agents make coverage decisions?
No. Use them to gather, summarise and draft; licensed staff should make determinations and approve customer communication.
Can it read claim documents?
You pass document text as context today; the files endpoint is coming soon. Check the docs for current status before planning a pipeline around uploads.
How do we evidence what the agent did?
Run every call with a named key and retain request, tool and approval logs so an auditor can reconstruct the flow.