With Recansio, and without.

Two tests on real legal questions, each with its method and its limits.

Tokens, time and precision.
24 questions, 16 November 2026.

Questions written by a practising lawyer, with the lawyer's answers. One model (Claude Opus 5.5) answered them twice: once with web search only, once with Recansio only. Three AI graders scored both answers against the lawyer's, in random order and without being told which side wrote which. They saw the text of every paragraph an answer linked to.

14×

fewer tokens than pasting the acts for each question

With Recansio
9,500 a question
Uploading the acts
131,000
Web search, as counted
7,500
3.4×

faster than searching the web

With Recansio
18 s a question
Web search
62 s
All 24
7.3 min / 25.0 min
24 of 24

answers rest on the exact paragraph

With Recansio
20 correct, 24 exact
Web search
21 correct, 1 exact
Question, by areaWith RecansioIts linkWeb searchIts link
1. Anti-money-laundering · asked in Uzbekcorrectexact paragraphcorrectnear it
2. Anti-money-laundering · asked in Englishcorrectexact paragraphcorrectnear it
3. Banking · asked in Russianpartlyexact paragraphcorrectnear it
4. Banking · asked in Russianpartlyexact paragraphcorrectnear it
5. Banking · asked in Uzbekcorrectexact paragraphcorrectno paragraph
6. Civil law · asked in Russiancorrectexact paragraphcorrectnear it
7. Civil law · asked in Russiancorrectexact paragraphcorrectnear it
8. Civil law · asked in Russiancorrectexact paragraphcorrectnear it
9. Criminal and administrative liability · asked in Englishcorrectexact paragraphpartlynear it
10. Criminal and administrative liability · asked in Russiancorrectexact paragraphpartlynear it
11. Criminal and administrative liability · asked in Uzbekcorrectexact paragraphcorrectno paragraph
12. Criminal and administrative liability · asked in Uzbekcorrectexact paragraphcorrectno paragraph
13. Criminal and administrative liability · asked in Uzbekcorrectexact paragraphcorrectno paragraph
14. Insurance and capital market · asked in Uzbekcorrectexact paragraphwrongnear it
15. Insurance and capital market · asked in Russianwrongexact paragraphcorrectnear it
16. Lending · asked in Uzbekcorrectexact paragraphcorrectnear it
17. Lending · asked in Uzbekpartlyexact paragraphcorrectno paragraph
18. Labour and tax · asked in Uzbekcorrectexact paragraphcorrectnear it
19. Labour and tax · asked in Uzbekcorrectexact paragraphcorrectnear it
20. Labour and tax · asked in Uzbekcorrectexact paragraphcorrectnear it
21. Payments and currency · asked in Russiancorrectexact paragraphcorrectnear it
22. Payments and currency · asked in Russiancorrectexact paragraphcorrectexact paragraph
23. Payments and currency · asked in Uzbekcorrectexact paragraphcorrectnear it
24. Public law · asked in Uzbekcorrectexact paragraphcorrectnear it
How tokens were counted
The whole run of the assistant for each batch of six questions, minus what an empty run costs (35,875 tokens), divided by six. For uploading: the Russian text of the acts each answer rests on, read from the base, at 0.454 tokens a character, which we measured by having the same model read 119,824 characters of the Labour Code. The middle value across the 24 questions is shown; 11 of them rest on two acts or more. Uploading all 29 acts once would take 2.9 million tokens.
What the web-search count leaves out
The pages the assistant opens are read by a separate tool, and those tokens are not in its count. So its own count (7,500 a question) is close to Recansio's, while the work took 3.4 times as long.
What "exact" means
At least one paragraph the answer links to itself states the deciding rule. "Near" means a link to the right act or a neighbouring part. The graders saw the text of every linked paragraph, read from the base.
Limits
24 questions: a difference of one or two answers is noise. The graders are models, not lawyers; the three agreed on 24 of 24 verdicts for Recansio and 22 for web search. If all your questions sit in one act, you upload it once per conversation, not once per question. The question texts stay unpublished so the set can be used blind again.

An earlier test: five assistants, 30 and 42 questions.

A strong assistant with web search can already answer correctly: ChatGPT got 30 of 30 on its own, and 28 with Recansio. What no assistant did on its own was show where the law says so. Of 180 answers given without Recansio (Claude 12, the other four 42 each), none linked to the exact paragraph. With Recansio, 118 of 126 did. That is what Recansio is for: an answer you can check.

Exact paragraph cited

Answers that link to the paragraph of the law, not to the act or a news site. Claude on its own was measured on 12 answers, the others on 42.

Correct answers

30 client questions in three languages.

Time

From sending the questions to the final answer. Less is better.

A free model on a laptop

Qwen 27B, offline, 64 GB of memory.

On its ownWith Recansio
Correct, 30 client questions524
Correct, 12 hard questions14
Wrong, 30 client questions152
Cite the exact paragraph, 42 answers040

In practice

Product risk assessment in a bank. Seen in use, not a controlled test.

5–10 working daysper specialist, before
10–20 minutesall departments, with Recansio

Method and limits

Questions
30 that clients asked, and 12 hard ones: past dates, appendices, licences, sanctions.
Grading
Three graders, each blind to which system wrote the answer. Majority verdict.
Limits
Small sets, one run each. One wrong answer moves a score by 3 to 8 points.

Full methodology

Try it on your own questions.

Test access opens as soon as you sign up: 100 calls over 7 days, no payment details.