---
title: "أفضل ذكاء اصطناعي للعمل القانوني في 2026: تحديث أكتوبر + 3,000 إجابة مقيّمة"
url: https://www.haqq.ai/ar/blog/best-ai-for-legal-work-benchmark
locale: ar
type: blog
description: "أكتوبر 2026: Claude Opus 5.5 وGPT-6 Astra يتصدران أحدث اختباراتنا القانونية. إضافة إلى 3,000 إجابة مقيّمة من 10 نماذج، 24% منها تستشهد بقانون لا يدعم ما تقوله."
category: "ai-legal-tech"
keywords: "best AI for legal work, best LLM for lawyers 2026, legal AI benchmark 2026, Claude Opus 5.5 legal, GPT-6 Astra legal, Claude Fable 5.1 legal, Gemini 3.1 Pro legal, DeepSeek V4 Pro legal, AI legal citation hallucination, most accurate AI for legal research, legal AI leaderboard, Claude vs GPT vs Gemini legal, frontier model legal benchmark"
published: 2026-06-05
updated: 2026-10-20
english_version: https://www.haqq.ai/blog/best-ai-for-legal-work-benchmark
source: src/data/blogPosts.ts
generator: scripts/generate-page-markdown.ts
---

# أفضل ذكاء اصطناعي للعمل القانوني في 2026: تحديث أكتوبر + 3,000 إجابة مقيّمة

أكتوبر 2026: Claude Opus 5.5 وGPT-6 Astra يتصدران أحدث اختباراتنا القانونية. إضافة إلى 3,000 إجابة مقيّمة من 10 نماذج، 24% منها تستشهد بقانون لا يدعم ما تقوله.

> **Note:** In short: as of October 2026, **Claude Opus 5.5** and **GPT-6 Astra** are the strongest frontier models we have tested for legal work. They were the top pair on our 30 September MENA trap test (3.30 and 3.00 of 4), and neither invented a Dubai court judgment when asked for one. Opus 5.5 matched Fable 5.1 at half the cost (23 September).

## Key facts

- On a five-question UAE, Saudi and Lebanese trap test, Claude Opus 5.5 scored 3.30 of 4 and GPT-6 Astra 3.00, against 2.20 for DeepSeek V4 Pro and 2.10 for Gemini 3.1 Pro, with hints removed (HAQQ run, 30 Sep 2026, 80 answers, three blind judges).
- Asked for Dubai Court of Cassation judgments supporting a non-existent article, Gemini 3.1 Pro and DeepSeek V4 Pro named 16 across eight answers; none matched a public source as cited (HAQQ run, 30 Sep 2026).
- Claude Opus 5.5, Opus 5 and Fable 5.1 finished within four verdicts of each other across 120 blind pairwise verdicts; Opus 5.5 cost $0.3647 per answer against $0.7249 for Fable 5.1 (HAQQ run, OpenRouter list prices, 23 Sep 2026).
- GPT-6 Astra led legal substance at 17.63 of 20 against Claude Opus 5 and GPT-5.6 Luna Pro, and all three got one Saudi or Egyptian rule wrong while citing the right article (HAQQ run, 41 prompts, 10 Sep 2026).
- Vals AI's Legal Research Bench placed Claude Opus 5.5 at 50.48%, 4th of 68 models (Vals AI, read 23 Sep 2026).
- In our largest run, 24% of 3,000 answers from 10 frontier models cited or applied law that does not say what the model claimed (HAQQ 300-task run, published 5 Jun 2026).
- Stanford RegLab measured hallucinations in about 33% of Westlaw AI-Assisted Research answers and more than 17% of Lexis+ AI answers (Magesh et al., 2024).

> **Note:** **Disclosure.** HAQQ builds legal AI and ran every HAQQ benchmark on this page. The September runs did not include HAQQ. The 50-point platform leaderboard further down does include HAQQ, and HAQQ ranks first on it. Read our numbers as a vendor's numbers: the method for each run is described, and the September runs publish their prompts and answers.

## Which AI is best for legal work as of October 2026?

Three runs in September 2026 replaced the June leaderboard for the frontier-model question. Each tests something different, so read them together.

| Model | Vendor | Run | Run date | Result |
| --- | --- | --- | --- | --- |
| Claude Opus 5.5 | Anthropic | Five MENA trap questions | 30 Sep 2026 | 3.30 / 4, named 0 invented judgments |
| GPT-6 Astra | OpenAI | Five MENA trap questions | 30 Sep 2026 | 3.00 / 4, named 0 invented judgments |
| DeepSeek V4 Pro | DeepSeek | Five MENA trap questions | 30 Sep 2026 | 2.20 / 4, named 8 unverifiable judgments |
| Gemini 3.1 Pro | Google | Five MENA trap questions | 30 Sep 2026 | 2.10 / 4, named 8 unverifiable judgments |
| Claude Opus 5 | Anthropic | Ten drafting and advice prompts, pairwise | 23 Sep 2026 | 39 of 80 verdicts, $0.4362 per answer |
| Claude Fable 5.1 | Anthropic | Ten drafting and advice prompts, pairwise | 23 Sep 2026 | 36 of 80 verdicts, $0.7249 per answer |
| Claude Opus 5.5 | Anthropic | Ten drafting and advice prompts, pairwise | 23 Sep 2026 | 35 of 80 verdicts, $0.3647 per answer |
| GPT-6 Astra | OpenAI | 41 prompts, 20 practice areas | 10 Sep 2026 | 17.63 / 20 substance, $0.2622 per answer |
| GPT-5.6 Luna Pro | OpenAI | 41 prompts, 20 practice areas | 10 Sep 2026 | 30.93 / 35 composite (with speed), $0.0231 per answer |
| Claude Opus 5 | Anthropic | 41 prompts, 20 practice areas | 10 Sep 2026 | 28.13 / 35 composite, $0.4566 per answer |
| Claude Opus 4.8 | Anthropic | 300 commercial tasks (superseded) | Jun 2026 | 30.02 / 35, won 130 of 300 tasks |

### October 2026 update: which result is current

What the September runs agree on: the quality gap between current frontier models on legal tasks is small, and the useful differences are in how a model fails (inventing authority, misreading a local rule) and what it costs per answer. Test the shortlist on your own matters before you pick.

## Which AI is most accurate for legal research?

On the measure that matters most in research, whether the authority exists and says what the model claims, Claude Opus 5.5 and GPT-6 Astra behaved best in our latest run: zero invented judgments in the eight answers between them. On Vals AI's independent Legal Research Bench, Opus 5.5 scored 50.48%, 4th of 68 models when we read it on 23 September 2026. No model we have tested is accurate enough to cite without checking. In September, all three models in the Astra run gave the right article number attached to the wrong rule on a Saudi or Egyptian question, which a citation checker would pass.

That is the gap a research tool has to close above the model: retrieve the primary source, check the citation against it, and say when it cannot find one. It is how we built [HAQQ legal research](https://www.haqq.ai/product/legal-research).

## Does a bigger or newer model hallucinate less?

Newer models hallucinate less often, and they still do it. Between June and September the pattern changed: in June, every one of the 10 models fabricated or misapplied a citation somewhere in 300 tasks, and 24% of all answers had a citation problem. In September, the newest Anthropic and OpenAI flagships declined to invent case law when asked for it. Gemini 3.1 Pro and DeepSeek V4 Pro still produced case numbers on demand, a different pair each time we asked. Model size alone did not predict the behaviour; the two that refused were also the two most expensive per answer in that run.

## Earlier run: NY litigation strategy (Jan 2026)

In January 2026 we gave seven systems one prompt: "Prepare litigation strategy under New York law for unpaid invoice of $250,000." We scored each answer out of 10 on six criteria. Strong answers used CPLR 3213 (summary judgment in lieu of complaint), pleaded breach of contract and account stated, and planned post-judgment collection (restraining notices, bank levies, turnover orders). Weaker answers stopped at filing suit. It was a single prompt scored by HAQQ, with HAQQ among the entrants, and every general model in it has since been replaced, so read it as a record of how answers differed, not as a ranking.

| Model | Legal accuracy | Procedural depth | Strategy | Authorities | Structure | Client-ready |
| --- | --- | --- | --- | --- | --- | --- |
| HAQQ | 9.5 | 9.5 | 9 | 9 | 9.5 | 9.5 |
| Claude Opus 4.6 | 9 | 8.5 | 9 | 9 | 9 | 8.5 |
| GPT-5.2 | 8.5 | 8.5 | 9 | 7 | 8.5 | 8.5 |
| Gemini 3.1 Pro | 8 | 8 | 8 | 7.5 | 8 | 8 |
| Grok 4.1 | 7.5 | 7 | 7 | 7 | 7.5 | 7 |
| Mistral Large 3 | 7 | 7 | 6.5 | 6 | 7 | 6.5 |
| Perplexity Sonar | 6 | 6 | 6 | 6 | 6.5 | 6 |

## How did we run the June 300-task benchmark?

We wrote **300 original legal tasks** with named parties, amounts, dates and governing statutes: draft this clause, redline this provision, structure this transaction, analyse this conflict of laws. They span 51 practice areas and more than 20 jurisdictions, including US federal law and Delaware, California, New York and Texas, the UK, the EU, the UAE and DIFC, Saudi Arabia, Lebanon, Egypt, Qatar, Singapore, Australia, Canada, India, Brazil, Nigeria, OHADA, Japan, Germany, France and Switzerland, plus the Hague Conventions and Cayman, BVI and Bermuda.

Difficulty was weighted hard: 114 tasks at level 5 of 5, 108 at level 4, 22 at level 3. Every task went to all 10 models with an identical system prompt at temperature 0, capped at 6,000 output tokens. Each answer was scored on five dimensions:

- **Quality (1 to 10)**: depth, completeness, actionability.
- **Accuracy (1 to 10)**: legal correctness, minus 5 for hallucinated case law and minus 3 for the wrong jurisdiction.
- **Speed (1 to 5)**: measured latency.
- **Style (1 to 5)**: professional structure.
- **Creativity (1 to 5)**: non-obvious risks and cross-jurisdiction issues spotted.

Claude Sonnet 4.6 scored quality, accuracy, style and creativity against a fixed rubric; speed came from real latency. The caveats section covers judge bias.

## What did the June 2026 leaderboard show?

- **Claude Opus 4.8** placed first in 130 of 300 tasks (nearly double any other model), top three in 265, with the highest quality (8.9) and creativity scores. It was slow (60.8 s) and among the most expensive ($0.069 per task).
- **Grok 4.3** placed second (28.98) in 8.8 seconds at $0.003 per task, the strongest result for speed and cost, and won the most environmental/ESG, IP and edge-case tasks.
- **GPT-5.5** posted the highest accuracy (8.41) and the lowest hallucinated-citation rate (3%), yet placed fifth on total and won one task outright. It was the slowest (134 s) and the most expensive ($0.082).
- **o3** won 66 tasks outright and still ranked eighth, with a 32% hallucination rate and accuracy of 5.89.
- **Mistral Large** misapplied or invented citations in 64% of answers; **Llama 4 Maverick** came last (20.01).

## How often do AI models cite law that does not support the answer?

In June, **24% of all 3,000 answers** cited or applied law that does not say what the model claimed: invented cases, misapplied statutes, the right doctrine attached to the wrong jurisdiction. The judge flagged specific, checkable errors. One example per model:

- **Claude Opus 4.8:** *Computer Associates / Gemstar cited as gun-jumping precedents; not accurate characterizations.*
- **GPT-5.5:** *Hitz citation appears fabricated; 616 B.R. 374 unverifiable.*
- **Gemini 3.1 Pro:** *Halpin v. Riverstone citation unverifiable; likely hallucinated Delaware case.*
- **Grok 4.3:** *CBS v. Ziff-Davis cited inaccurately; case concerns warranty reliance, not disclosure.*
- **Claude Sonnet 4.6:** *Schrier cite unverifiable; Biotronik case real but misapplied.*
- **o3:** *MSA Technology v Antec and GB Gas citations appear fabricated or misrepresented.*
- **Qwen3.7 Max:** *Specht v. Netscape misapplied; not a goods/services UCC case.*
- **DeepSeek V3.2:** *Crosstown Music citation unverifiable / misapplied.*
- **Mistral Large:** *N.Y. Gen. Oblig. Law §5-328 and UCC §2-515 citations are dubious / misapplied.*
- **Llama 4 Maverick:** *Cavendish case misapplied; unrelated to pricing clauses.*

Every model, including the leaders, made at least one such error, and the top accuracy score was 8.41 of 10 (GPT-5.5). The research tools lawyers already pay for show the same problem: Stanford RegLab's 2024 study measured hallucinations in about a third of Westlaw AI-Assisted Research answers and more than one in six for Lexis+ AI.

## Does one model win every practice area?

No. Across the 51 practice areas in June, Claude Opus 4.8 led 30, Grok 4.3 led 13, o3 led 6, and Gemini 3.1 Pro and Claude Sonnet 4.6 led one each.

- **Opus 4.8** led the doctrinal core: M&A (30.6), regulatory compliance (30.5), securities, white collar, tax, banking, data privacy, employment and international trade.
- **Grok 4.3** led IP and technology (30.2), environmental and ESG (30.8), consumer protection, and compliance and due diligence (31.4).
- **o3** led legal operations (31.1), securities and finance (30.9), government contracts and procurement (31.1) and cross-border real estate.
- **Gemini 3.1 Pro** posted the top score in privacy and data protection (31.4).

The September Astra run found the same shape at a smaller scale: Astra ahead in most practice areas, Claude Opus 5 ahead in insolvency and ESG. A legal product that routes each matter to the model most likely to get it right, then checks the answer, gets more out of the field than one that bets on a single model.

## How much does each model cost per legal task?

In June, cost per task spanned a 90x range, from $0.0009 (DeepSeek V3.2) to $0.082 (GPT-5.5), and latency ran from 7.7 seconds (Llama 4 Maverick) to 134 seconds (GPT-5.5). The September runs, priced at OpenRouter list prices on the run date, give the current picture:

- Claude Opus 5.5: $0.3647 per answer at high reasoning effort, against $0.7249 for Fable 5.1 and $0.4362 for Opus 5 (23 Sep 2026).
- GPT-6 Astra: $0.2622 per answer, 1.7x cheaper than Claude Opus 5 because it writes about 3.5x fewer tokens, and 11x more than GPT-5.6 Luna Pro at $0.0231 (10 Sep 2026).
- On the trap test at default settings, the 20 answers per model cost USD 4.04 for Opus 5.5, 2.81 for Astra, 0.92 for Gemini 3.1 Pro and 0.88 for DeepSeek V4 Pro (30 Sep 2026).

For a high-volume product, the model with the top quality score can be the wrong choice once cost and latency are counted. Astra's raised reasoning effort is a case in point: in September, moving it from low to high multiplied cost by 1.5x and latency by 1.8x and moved judged quality by minus 0.17 points.

## How do legal AI platforms compare with frontier models?

We keep a second leaderboard: a 50-task evaluation on a 50-point rubric across 11 task categories that adds the legal platforms firms shortlist (Harvey, CoCounsel, Legora, Spellbook, LexisNexis +AI, Clio Duo) next to frontier models. It is the leaderboard on our compare page, and HAQQ ranks first on it (see the disclosure above). Ranked by average score:

| Rank | Model or product | Average /50 |
| --- | --- | --- |
| 1 | HAQQ (Justinian) | 46.8 |
| 2 | Claude Fable 5 | 42.7 |
| 3 | Claude Opus 4.8 | 40.8 |
| 4 | Mike OS | 39.1 |
| 5 | Harvey Tenet | 37.9 |
| 6 | DeepSeek V4 Pro | 36.8 |
| 7 | Thomson Reuters CoCounsel | 36.6 |
| 8 | Legora | 35.9 |
| 9 | GPT-5.6 Sol Pro | 35.3 |
| 10 | Claude + legal plugins | 33.9 |
| 11 | Gemini 3.1 Pro | 32.9 |
| 11 | Spellbook | 32.9 |
| 13 | LexisNexis +AI | 32.0 |
| 14 | Grok 4.5 | 30.7 |
| 15 | Perplexity Sonar | 27.2 |
| 16 | Clio Duo | 25.6 |
| 17 | Llama 4 Maverick | 23.8 |
| 18 | Mistral 3 | 22.4 |
| 19 | Qwen3.7 Plus | 18.7 |

Two readings stand out. The legal platforms cluster in the middle: Harvey Tenet (37.9), CoCounsel (36.6), Legora (35.9) and LexisNexis +AI (32.0) all sit below the strongest frontier models on standalone answer quality, so what a firm pays a platform for is mainly workflow, integrations and security. And the spread is wide, from 46.8 to 18.7, so the choice of tool changes the output a lawyer starts from. A per-task score cannot see firm-wide deployment, certifications or document-scale workflows, which are real strengths of the larger platforms.

The platform leaderboard has not yet been re-run on Opus 5.5, Fable 5.1 or GPT-6 Astra; the September runs above are the current source for those models.

## Which provider leads across every run?

Collapsing models to provider and merging the June run with every earlier run of the same commercial benchmark (0 to 100 index, weighted by number of evaluations):

The top providers sit close together. What you build around the model (retrieval, citation checks, jurisdiction rules) moves the result more than which of them you pick.

## Can I trust these benchmarks?

- **Output caps.** Our first June pass capped answers at 1,200 tokens, which starved reasoning models and clipped verbose ones. We discarded it and re-ran at 6,000. Most models gained 1.5 to 2.5 points; Mistral and Llama got worse, because longer answers exposed more bad citations. In September a 32,000-token cap truncated two Opus answers on one prompt; we re-ran at 64,000 and now abort any run with a truncated answer.
- **Judges.** June used one judge, Claude Sonnet 4.6, so self-preference is a fair question; Grok 4.3 placing second and GPT-5.5 posting the top accuracy argue against a strong home bias, but a second judge would be better. The September runs used two non-Anthropic judges (23 Sep) and three judges from three model families with the median taken (30 Sep). On 23 September the two judges picked different winners from the same answers.
- **Sample size.** The September runs used 10, 41 and 5 prompts. That is enough to see a large difference and too small to rank models that are close. Read effect sizes, not decimals.
- **Single runs.** June was a single run at temperature 0. In September, repeated answers moved by 0.3 points on average and by a full point or more in 11 of 40 pairs.
- **Human review.** No practising lawyer graded the September answers, and Arabic drafting was judged by models only.

## How should a firm choose a model today?

- Shortlist Claude Opus 5.5 and GPT-6 Astra for research and advice where invented authority is the main risk; both declined to fabricate judgments in our latest test.
- Price the shortlist on your own volume. Opus 5.5 is about half the cost of Fable 5.1 per answer for quality we could not tell apart; GPT-5.6 Luna Pro is far cheaper than Astra for general Q&A where a point of quality is not worth 11x.
- Test on your own jurisdictions. Every model in the Astra run misread one Saudi or Egyptian rule while citing the right article.

## Sources

- [HAQQ, Claude Opus 5.5 legal benchmark (23 Sep 2026)](https://www.haqq.ai/blog/claude-opus-5-5-legal-benchmark)
- [HAQQ, GPT-6 Astra on legal work (10 Sep 2026 run)](https://www.haqq.ai/blog/gpt-6-astra-legal-work)
- [HAQQ, How to test a legal AI: five trap questions (30 Sep 2026 run)](https://www.haqq.ai/blog/how-to-test-legal-ai)
- [Vals AI, Legal Research Bench](https://www.vals.ai/benchmarks/legal_research)
- [Anthropic, Introducing Claude Opus 5.5](https://www.anthropic.com/claude-opus-5-5)
- [Magesh et al., Stanford RegLab study of AI legal research tools (2024)](https://arxiv.org/abs/2405.20362)

## Related reading

- [the court-filing hallucination audit](https://www.haqq.ai/blog/ai-legal-hallucination-audit)
- [nine AI systems on a Lebanese waqf matter](https://www.haqq.ai/blog/lebanese-family-waqf-legal-ai-benchmark)
- [when an LLM judge rewards fabrication](https://www.haqq.ai/blog/jev-legal-ai-hallucination-detection)
- [the scope and method of our MENA legal AI benchmark](https://www.haqq.ai/blog/mena-legal-ai-benchmark-scope-and-method)
- [why LLMs fabricate, explained for lawyers](https://www.haqq.ai/blog/lawyers-guide-to-large-language-models)

## الأسئلة الشائعة

### ما أفضل ذكاء اصطناعي للعمل القانوني في 2026؟

حتى أكتوبر 2026، يُعدّ Claude Opus 5.5 وGPT-6 Astra أقوى النماذج الرائدة في اختباراتنا. ففي اختبار من خمسة أسئلة فخّية عن قوانين الإمارات والسعودية ولبنان أُجري في 30 سبتمبر 2026 حصلا على 3.30 و3.00 من 4، متقدمَين على DeepSeek V4 Pro (2.20) وGemini 3.1 Pro (2.10)، ولم يختلق أي منهما حكماً قضائياً حين طُلب منه ذلك. والفارق بينهما ضمن التذبذب بين التشغيلات، فليكن السعر وقضاياك أنت هما الفيصل.

### ما أدق ذكاء اصطناعي في البحث القانوني؟

في تشغيلنا بتاريخ 30 سبتمبر 2026، امتنع Claude Opus 5.5 وGPT-6 Astra عن ذكر أحكام لمحكمة تمييز دبي بشأن مادة غير موجودة، بينما ذكر Gemini 3.1 Pro وDeepSeek V4 Pro ستة عشر حكماً لا يطابق أي منها مصدراً عاماً كما استُشهد به. ووضع معيار Legal Research Bench المستقل من Vals AI نموذج Opus 5.5 في المرتبة الرابعة من 68 نموذجاً بنسبة 50.48% (بتاريخ 23 سبتمبر 2026). ولا يوجد نموذج دقيق بما يكفي للاستشهاد به دون التحقق من المصدر.

### هل يهلوس نموذج الذكاء الاصطناعي الأكبر أو الأحدث أقل في الأسئلة القانونية؟

النماذج الرائدة الأحدث تهلوس أقل، لكنها لا تزال تهلوس. ففي تشغيلنا في يونيو 2026 على 3,000 إجابة، استشهدت 24% منها بقانون لا يدعم الادعاء أو طبّقته، وأخطأ كل نموذج من النماذج العشرة في موضع ما. وفي سبتمبر 2026 رفضت أحدث نماذج Anthropic وOpenAI اختلاق سوابق قضائية، لكن النماذج الثلاثة جميعها في تشغيلنا الخاص بـ GPT-6 Astra أخطأت في بيان قاعدة سعودية أو مصرية واحدة مع استشهادها بالمادة الصحيحة.

### هل Claude Opus 5.5 أفضل من Claude Fable 5.1 في العمل القانوني؟

لم نستطع التمييز بينها من حيث الجودة. فعبر 120 حكماً أعمى من حَكَمين ليسا من Anthropic في 23 سبتمبر 2026، جاءت نماذج Opus 5.5 وOpus 5 وFable 5.1 ضمن أربعة أحكام من بعضها، وشكّلت حلقة لا ترتيباً. وكلّف Opus 5.5 مبلغ 0.3647 دولار للإجابة مقابل 0.7249 دولار لـ Fable 5.1، وكان الأسرع بين الثلاثة.

### كم تكلفة تمرير سؤال قانوني عبر نموذج رائد؟

في تشغيلاتنا في سبتمبر 2026 بأسعار OpenRouter المعلنة، كلّف GPT-5.6 Luna Pro نحو 0.02 دولار للإجابة، وGPT-6 Astra نحو 0.26 دولار، وClaude Opus 5.5 نحو 0.36 دولار، وClaude Opus 5 بين 0.44 و0.46 دولار، وClaude Fable 5.1 نحو 0.72 دولار. ويحدّد طول الإجابة الفاتورة بقدر ما يحدّدها سعر الرمز: فـ Astra يتقاضى ضعف سعر Opus 5 للرمز، ومع ذلك تقل تكلفته للإجابة بـ1.7 مرة لأنه يكتب رموزاً أقل بنحو 3.5 مرة.

### أين تقع منصات الذكاء الاصطناعي القانونية مثل Harvey مقارنةً بالنماذج الرائدة؟

على معيار المهام من HAQQ بمقياس 50 نقطة عبر 11 فئة و19 مشاركاً، يبلغ متوسط Harvey Tenet 37.9/50، متقدماً على CoCounsel (36.6) وLegora (35.9) وLexisNexis +AI (32.0)، ومتأخراً عن Claude Fable 5 (42.7) وClaude Opus 4.8 (40.8). وأفضل ترتيب لـ Harvey في أي فئة منفردة هو المرتبة 2. وقد أعدّت HAQQ هذا المعيار وتحتل المرتبة الأولى فيه بمتوسط 46.8/50، فتعامل معه بوصفه لوحة صدارة من إعداد مورّد وتحقق من المنهجية.

### ما أفضل نموذج Claude للعمل القانوني؟

على معيار HAQQ القانوني بمقياس 50 نقطة (النتائج حتى 20 أكتوبر 2026)، بلغ متوسط Claude Fable 5 42.7/50 وClaude Opus 4.8 40.8، وكلاهما متقدم على Harvey (37.9). وسجّل Claude مع الإضافات القانونية 33.9، أي أقل من Claude العادي. يناسب Claude العادي التحليل والمسودات الأولى والمستندات الطويلة، لكنه ليس نظام سجلات ولا أداة للتحقق من الاستشهادات. وقد أجرت HAQQ هذا المعيار وتحتل المرتبة الأولى فيه، فأعد الاختبار على قضاياك أنت.

### أين أجد نتائج معايير مستقلة لأدوات الذكاء الاصطناعي القانونية؟

قاس باحثو Stanford RegLab (Magesh وآخرون، 2024) بشكل مستقل معدلات هلوسة بلغت نحو 17% لـ Lexis+ AI و33% لـ Westlaw AI-Assisted Research. أما معيار HAQQ بمقياس 50 نقطة لـ 19 أداة ونموذجاً فينشر نتيجة كل فئة على haqq.ai/compare-us، لكنه ليس مستقلاً: فقد أجرته HAQQ، وتحتل المرتبة الأولى فيه، ولم يدقّقه أي طرف ثالث.

---

Canonical page: https://www.haqq.ai/ar/blog/best-ai-for-legal-work-benchmark
Whole-site context for agents: https://www.haqq.ai/llms.txt · https://www.haqq.ai/agents.md

Generated at build time from the site's own content data. Do not edit by hand.
