Prompt size at turn 12
50× smallerCrystal footprint vs raw-accumulating at the final turn. Crystal: 1,093 tokens. Raw accumulating: 54,324 tokens — 50× larger and growing every call.
Insurance Carriers plan
Carriers need policy memory before claim judgment.
Buy an insurance-carriers library built around shared coverage, claims, and regulatory context before adding private program material.
Start free 7-day trial
Free for 7 days. No card required. 30,000 credits and the full Insurance Carriers library. Sign in with Google or Apple using your work email. Keep it after the trial at $4,500/mo.
Multi-turn benchmark
Compact through 12 turns. No degradation detected.
Three arms, same underlying model, same question set — no context, raw-accumulating, hosted crystal. At turn 12 the crystal sent 1,093 tokens; the raw arm sent 54,324 — 50× more, compounding every call.
Instruction retention
All arms retainedA formatting rule planted at turn 1 was checked deterministically every subsequent turn. No-context retained it through all checkable turns. Crystal: 1 / 1 checkable turns held.
Answer quality — 12 turns
No degradation detectedAll three arms produced substantive answers through turn 12 with no meaningful quality decline. The separating factors are prompt size and instruction retention, not raw answer quality.
How this was run. 12 turns, Insurance Carriers vertical, The same model held constant across all three arms. Raw-accumulating arm: source corpus injected at turn 1, then full conversation history re-sent on every subsequent turn — the window grows every call.
Free trial
Try it free for 7 days.
Free for 7 days. No card required. Keep it after the trial from $4,500/mo. It starts with a shared policy, claims and regulatory reference; private program material is connected when you decide.
Your 7 free days
$0
Free for 7 days. No card required. 30,000 credits and the full Insurance Carriers library. Sign in with Google or Apple using your work email.
If you keep it
$4,500/mo · about $150 a day
6,000,000 credits/mo. Up to 20M raw-token-equivalent shared insurance reference coverage. Choose it when your trial ends; nothing is charged until you do.
Raw RAG can inflate every prompt with retrieved text. The crystal keeps shared operating judgment compact and reusable, so teams buy a monthly memory layer instead of repeatedly paying for larger, noisier context windows.
Context crystals
Representative coverage.
Each plan includes protected access to prepared industry memory across the recurring work areas below. Your team sees the benefit in chat while the library's source construction remains private.
Coverage Language
A policy-language reference layer for separating coverage concepts, exclusions, endorsements, and common interpretive questions.
- Policy-form clause purpose mapping
- Coverage and exclusion language cues
- Endorsement relationship checkpoints
Claims Operations
A claims-context layer for triage language, documentation expectations, and recurring handling patterns.
- Claims intake issue classification
- Reserve and exposure vocabulary alignment
- Documentation sufficiency review cues
Regulatory Change
A public regulatory reference layer for bulletins, market conduct topics, and state-facing compliance language.
- State bulletin topic taxonomy
- Market conduct obligation cues
- Filing-language consistency review
Try the Insurance Carriers library free for 7 days.
Free for 7 days. No card required. 30,000 credits and the full Insurance Carriers library. Sign in with Google or Apple using your work email. Keep it after the trial at $4,500/mo, or talk to us about a rollout built around your team.
Use cases
What Insurance Carriers teams run on it.
Starting points to test during the free trial. Your professionals keep every judgment; the library supplies shared reference.
Coverage language comparison
Ask how a policy provision is typically read and where wording differences change the question, then route the real determination to an adjuster or coverage counsel.
Claims triage briefs
Summarize a claim file against shared claims-operations reference so the handler starts with the right questions and the right checklist.
Regulatory change watch
Run the same question against each new bulletin or rule change to get a consistent first read for compliance and product teams.
Connect it to your work
Build it into the way your team already runs.
Two HTTPS calls. Create a handle for your own authorized material with /v1/compress, then ask questions with /v1/context/run. Call it from a scheduled job, a webhook or an internal tool, and post the answer where your team already works. See the API docs.
Shared claims and compliance workspace
One workspace for claims, underwriting support and compliance, all asking against the same prepared insurance reference.
Scheduled regulatory digests
Call /v1/context/run each morning with fixed questions per line of business and post the answers to compliance and product channels.
Claims-system and inbox triggers
When a new claim, endorsement request or regulator notice arrives in your claims system or shared inbox, trigger a run call that returns a structured first-read for the human owner.
FAQ
Common questions.
Read the published evaluation or the guide to source-grounded answers.
Can this decide a claim?
No. It provides reference context and consistent first reads. Coverage determinations stay with your claims professionals and counsel.
How does a prompt-size comparison apply to claims work?
Claim files are long. In our published evaluation the peak prompt over a 12-turn thread was about 1,093 tokens with the hosted library versus about 54,324 with raw accumulating context, so each later turn re-reads far less.
Can we use our own policy forms?
Yes, when you choose to. Authorized forms and guidelines are connected separately through /v1/compress and asked about with /v1/context/run.
What does the Insurance Carriers plan include?
Free for 7 days. No card required. 30,000 credits and the full Insurance Carriers library. Sign in with Google or Apple using your work email. Keep it after the trial at $4,500/mo with 6,000,000 credits per month.