Released · Delivered by EncevraLexplio v4
The biggest jump the standard tier has ever had. Every $9.99 report now runs on the v4 engine: a new analysis core delivered by Encevra AI, configured exclusively for Lexplio and not offered to the public anywhere else. It pairs a 1M-token context window with 128K-token output capacity and a +47% independent intelligence score over v3. The full measured gap follows below.
By the numbers
better at multi-step analysis
On long, multi-part contract reviews, the v4 engine completes the full chain of checks nearly seven times more often than v3.
higher overall intelligence score
Independent evaluators score the v4 engine 47% higher than the v3 engine across reasoning, accuracy and document work.
better professional-grade output
Measured on expert-level professional knowledge work: the same task class as reading a lease and assigning a defensible risk grade.
faster report generation
Reports come back noticeably quicker, and each analysis costs less to run; the savings are passed on by holding the price at $9.99.
v3 engine vs. v4 engine
Both engines were evaluated on identical task suites under the same scoring protocol: multi-step verification chains over long documents, exact span extraction, and constrained in-place edits. Both share a 1M-token context window and the same underlying cost structure. Everything below is the measured gap between the engine behind v3 and the one behind v4, expressed in the capabilities that drive a Lexplio report.
| Capability | v3 engine | v4 engine | Gain vs v3 |
|---|---|---|---|
| Multi-step contract reviewFollowing a full chain of checks across a long document without dropping steps | 10.3% | 70.6% | +60.3 pts · ~6.9× |
| Complex task resolutionWorking through layered, real-world problems to a correct answer | 63.2 | 81.3 | +18.1 pts · +28.6% |
| Professional knowledge workQuality of expert-level analysis and written deliverables | 1449 | 1844 | +27.3% |
| Structured document deliverablesProducing complete, well-organized reports from dense source material | 1359 | 1811 | +33.3% |
| Precise clause locatingPinpointing the exact passage that matters inside a long document | 34.1% | 55.5% | +21.4 pts · +62.8% |
| Deep multi-step reasoningHard reasoning problems that require tools and multiple inference steps | 54.9% | 64.5% | +9.6 pts · +17.5% |
| End-to-end task executionOperating real interfaces and completing tasks start to finish | 57.0% | 80.1% | +23.1 pts · +40.5% |
| Dense data readingExact reading of dense, structured figures and tables | 15.6% | 61.6% | +46.0 pts · ~3.9× |
| Scoped, precise editsMaking constrained, in-scope changes without touching anything else | 42.4% | 52.1% | +9.7 pts · +22.9% |
Units: "pts" is an absolute percentage-point change; "%" is relative change. Scores are normalized pass rates and Elo-style ratings from published evaluation suites for each engine generation. Both engines were evaluated at their default operating settings.
Independent cross-checks
Our own numbers aren't the only signal. Three independent evaluators scored both engine generations head to head; every one of them places the v4 engine ahead, by margins of +7.5% to +47%.
Independent intelligence index
38 → 56+18 pts (+47.4%)The v4 engine ranks #3 out of 216 models scored by this independent evaluator, and it costs less per task than the v3 engine at most effort levels.
Independent composite score
47.9 → 51.5+3.6 (+7.5%)Wins every shared head-to-head against the v3 engine and lifts every category: coding-style precision +18.3%, agentic multi-step work +26.5%, reasoning +5.3%.
Independent category composite
66.98 → 80.49+13.51 (+20.2%)Category averages jump across the board. The largest single swing is on visually structured inputs: nearly a 2× improvement on documents with complex formatting.
Speed & economics
+30%
faster report generation
The v4 engine produces output 30%+ faster than v3: quicker report turnaround at the same price.
−30%
cost per analysis
It resolves each review in fewer steps: up to 30% less compute per report, which is how the price stays at $9.99.
128K
max output length
Up from prior output limits: long documents get complete, untruncated reports.
One trade-off to note: at maximum reasoning effort, per-analysis cost can run higher than v3. The efficiency gains land at the default effort levels where Lexplio's standard analyses operate.
What those numbers mean for your report
| Report capability | The evidence | Practical effect |
|---|---|---|
| Long-document reading | 1M-token context on both engines; +27% on professional knowledge work | Better at connecting a clause on page 4 to a penalty on page 40 without losing the thread. |
| Exact clause extraction | Dense data reading 15.6% → 61.6%; precise clause locating 34.1% → 55.5% | Cleaner verbatim quotes and fewer mis-anchored citations in your report. |
| Multi-step reasoning | Deep reasoning 54.9% → 64.5%; complex task resolution 63.2 → 81.3 | Stronger cross-reference of obligations, exceptions and defined terms across sections. |
| Findings precision | Professional-grade output within 2 points of the most expensive flagship engine tier | Work quality statistically indistinguishable from engines that cost far more to run. |
v3 vs. v4, side by side
| Metric | v3 | v4 |
|---|---|---|
| Engine provider | Lexplio v3 | Delivered by Encevra |
| Availability | Standard | Exclusive to Lexplio |
| Multi-step contract review | 10.3% | 70.6% |
| Professional knowledge work | 1449 | 1844 |
| Independent intelligence score | 38 | 56 |
| Generation speed | Baseline | 30%+ faster |
| Cost per analysis | Baseline | Up to 30% lower |
| Context window | 1M tokens | 1M tokens |
| Report sections | 20 | 20 |
| HOA / CC&R mode | Yes | Yes, sharper |
| Price | $9.99 | $9.99 |
What changed
A new engine, delivered by Encevra
Lexplio v4 is delivered through our parent company, Encevra AI, and configured exclusively for Lexplio. The v4 engine: its model setup, contract instructions and report pipeline are not offered as a product anywhere else.
The biggest jump for standard reports
Every $9.99 standard analysis now runs on v4. Expect tighter red flags, more accurate quotes, better hidden-fee detection and more consistent grades across long leases, job offers and HOA documents.
Same report you know
The 20-section report, letter grade, risk score, HOA / CC&R mode and suggested edits are all still there. The engine underneath is what got smarter.
Nothing to change on your end
No new plan, no price change. Pro members and one-time buyers get v4 automatically, and Noetis Deep Research stays available for the most thorough review.
Zero-retention, unchanged
Your contract text is never stored. It is processed in memory and discarded the moment your report is generated.
Lexplio v4: common questions
What is Lexplio v4?
Lexplio v4 is the newest standard contract analysis engine from Lexplio, delivered by Encevra AI and configured exclusively for Lexplio. It is the biggest capability jump in Lexplio's history.
How much better is Lexplio v4 than v3?
v4 scores 47% higher on an independent overall intelligence index, performs nearly 7x better on multi-step contract review, locates clauses far more precisely, and returns reports more than 30% faster.
How much does a Lexplio v4 analysis cost?
A full v4 analysis costs $9.99, the same price as before. Every new visitor also gets a free preview showing the risk score and top three red flags.
What does a Lexplio v4 report include?
An A to F grade, a risk score out of 10, severity-ranked red flags with exact quotes from your contract, hidden fees, state-specific notes, and suggested edits.
Does Lexplio store my contract?
No. Contract text is processed in real time and discarded immediately after analysis. Lexplio is an educational tool and does not provide legal advice.
Run your contract through v4
Paste or upload your document and see what the new engine finds.
Start an analysis