# Legal Domain Fabrication and Factuality Analysis on JURIDICO-FCR 75 tasks Source: [Legal Domain Fabrication and Factuality Analysis on JURIDICO-FCR 75 tasks]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:35:26.640Z Evaluation of model performance on the JURIDICO-FCR 75 tasks dataset specifically for legal domain fabrication and factuality analysis, reporting average and worst-case scores for Fabrication Rate (FCR) and Factuality Failure Rate (FFR). A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Legal Domain Fabrication and Factuality Analysis | | Dataset (including any stated split) | JURIDICO-FCR 75 tasks | | Metric | Average FCR | | Unit | PERCENT | | Direction | Lower is better | | Results loaded | 36 of 36 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Average FCR](): Lower is better; unit: PERCENT. - [Average FFR](): Lower is better; unit: PERCENT. - [Worst Task FCR](): Lower is better; unit: PERCENT. - [Worst Task FFR](): Lower is better; unit: PERCENT. ## Recorded results | Method | Average FCR | Result date | Report paper | | --- | --- | --- | --- | | GPT-5.2 (gpt-5.2-2025-12-11) | 0.11 | Not recorded | 69730060b2be0ed85a78c834 | | Gemini 3.0 Pro (gemini-3.0-pro) | 0.12 | Not recorded | 69730060b2be0ed85a78c834 | | GPT-4.1 (gpt-4.1-2025-04-14) | 0.13 | Not recorded | 69730060b2be0ed85a78c834 | | DeepSeek V3.1 Terminus (deepseek-v3.1-terminus) | 0.13 | Not recorded | 69730060b2be0ed85a78c834 | | Kimi K2 Thinking (Kimi-K2-Thinking) | 0.14 | Not recorded | 69730060b2be0ed85a78c834 | | Claude 4.5 Sonnet (claude-sonnet-4-5...) | 0.14 | Not recorded | 69730060b2be0ed85a78c834 | | Qwen3 235B A22B Instruct (Qwen3-235B-A22B...) | 0.15 | Not recorded | 69730060b2be0ed85a78c834 | | Gemma 3 27B (gemma-3-27b-it) | 0.15 | Not recorded | 69730060b2be0ed85a78c834 | | Llama 4 Maverick (Llama-4-Maverick...) | 0.16 | Not recorded | 69730060b2be0ed85a78c834 | | Mistral Small 24B (Mistral-Small-24B...) | 0.17 | Not recorded | 69730060b2be0ed85a78c834 | | Gemini Flash Latest (gemini-flash-preview...) | 0.18 | Not recorded | 69730060b2be0ed85a78c834 | | Claude 4.1 Opus (claude-opus-4-1...) | 0.19 | Not recorded | 69730060b2be0ed85a78c834 | | GPT-5.2 (gpt-5.2-2025-12-11) | 4.2 | Not recorded | 69730060b2be0ed85a78c834 | | Gemini 3.0 Pro (gemini-3.0-pro) | 4.9 | Not recorded | 69730060b2be0ed85a78c834 | | GPT-4.1 (gpt-4.1-2025-04-14) | 5.3 | Not recorded | 69730060b2be0ed85a78c834 | | Kimi K2 Thinking (Kimi-K2-Thinking) | 6.8 | Not recorded | 69730060b2be0ed85a78c834 | | DeepSeek V3.1 Terminus (deepseek-v3.1-terminus) | 7.1 | Not recorded | 69730060b2be0ed85a78c834 | | Claude 4.5 Sonnet (claude-sonnet-4-5...) | 8.1 | Not recorded | 69730060b2be0ed85a78c834 | | Qwen3 235B A22B Instruct (Qwen3-235B-A22B...) | 8.9 | Not recorded | 69730060b2be0ed85a78c834 | | Llama 4 Maverick (Llama-4-Maverick...) | 9.5 | Not recorded | 69730060b2be0ed85a78c834 | | Gemma 3 27B (gemma-3-27b-it) | 10.8 | Not recorded | 69730060b2be0ed85a78c834 | | Mistral Small 24B (Mistral-Small-24B...) | 12 | Not recorded | 69730060b2be0ed85a78c834 | | Gemini Flash Latest (gemini-flash-preview...) | 13.2 | Not recorded | 69730060b2be0ed85a78c834 | | Claude 4.1 Opus (claude-opus-4-1...) | 14.8 | Not recorded | 69730060b2be0ed85a78c834 | | GPT-5.2 (gpt-5.2-2025-12-11) | 15.1 | Not recorded | 69730060b2be0ed85a78c834 | | Gemini 3.0 Pro (gemini-3.0-pro) | 16.8 | Not recorded | 69730060b2be0ed85a78c834 | | GPT-4.1 (gpt-4.1-2025-04-14) | 18.5 | Not recorded | 69730060b2be0ed85a78c834 | | Kimi K2 Thinking (Kimi-K2-Thinking) | 21.9 | Not recorded | 69730060b2be0ed85a78c834 | | DeepSeek V3.1 Terminus (deepseek-v3.1-terminus) | 22.4 | Not recorded | 69730060b2be0ed85a78c834 | | Qwen3 235B A22B Instruct (Qwen3-235B-A22B...) | 26.5 | Not recorded | 69730060b2be0ed85a78c834 | | Claude 4.5 Sonnet (claude-sonnet-4-5...) | 28.2 | Not recorded | 69730060b2be0ed85a78c834 | | Gemma 3 27B (gemma-3-27b-it) | 30.2 | Not recorded | 69730060b2be0ed85a78c834 | | Llama 4 Maverick (Llama-4-Maverick...) | 31 | Not recorded | 69730060b2be0ed85a78c834 | | Mistral Small 24B (Mistral-Small-24B...) | 34.5 | Not recorded | 69730060b2be0ed85a78c834 | | Gemini Flash Latest (gemini-flash-preview...) | 36.8 | Not recorded | 69730060b2be0ed85a78c834 | | Claude 4.1 Opus (claude-opus-4-1...) | 40 | Not recorded | 69730060b2be0ed85a78c834 | ### GPT-5.2 (gpt-5.2-2025-12-11) Average FCR: 0.11. Result ID: 699624b85698665ec4656c5d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### Gemini 3.0 Pro (gemini-3.0-pro) Average FCR: 0.12. Result ID: 699624b85698665ec4656c60. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### GPT-4.1 (gpt-4.1-2025-04-14) Average FCR: 0.13. Result ID: 699624b85698665ec4656c63. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### DeepSeek V3.1 Terminus (deepseek-v3.1-terminus) Average FCR: 0.13. Result ID: 699624b85698665ec4656c66. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### Kimi K2 Thinking (Kimi-K2-Thinking) Average FCR: 0.14. Result ID: 699624b85698665ec4656c69. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### Claude 4.5 Sonnet (claude-sonnet-4-5...) Average FCR: 0.14. Result ID: 699624b85698665ec4656c6f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### Qwen3 235B A22B Instruct (Qwen3-235B-A22B...) Average FCR: 0.15. Result ID: 699624b85698665ec4656c6c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### Gemma 3 27B (gemma-3-27b-it) Average FCR: 0.15. Result ID: 699624b85698665ec4656c75. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### Llama 4 Maverick (Llama-4-Maverick...) Average FCR: 0.16. Result ID: 699624b85698665ec4656c72. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### Mistral Small 24B (Mistral-Small-24B...) Average FCR: 0.17. Result ID: 699624b85698665ec4656c78. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### Gemini Flash Latest (gemini-flash-preview...) Average FCR: 0.18. Result ID: 699624b85698665ec4656c7b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### Claude 4.1 Opus (claude-opus-4-1...) Average FCR: 0.19. Result ID: 699624b85698665ec4656c7e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Advanced RAG"}. This record does not assert independent reproduction. ### GPT-5.2 (gpt-5.2-2025-12-11) Average FCR: 4.2. Result ID: 699624b85698665ec4656c5c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### Gemini 3.0 Pro (gemini-3.0-pro) Average FCR: 4.9. Result ID: 699624b85698665ec4656c5f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### GPT-4.1 (gpt-4.1-2025-04-14) Average FCR: 5.3. Result ID: 699624b85698665ec4656c62. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### Kimi K2 Thinking (Kimi-K2-Thinking) Average FCR: 6.8. Result ID: 699624b85698665ec4656c68. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### DeepSeek V3.1 Terminus (deepseek-v3.1-terminus) Average FCR: 7.1. Result ID: 699624b85698665ec4656c65. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### Claude 4.5 Sonnet (claude-sonnet-4-5...) Average FCR: 8.1. Result ID: 699624b85698665ec4656c6e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### Qwen3 235B A22B Instruct (Qwen3-235B-A22B...) Average FCR: 8.9. Result ID: 699624b85698665ec4656c6b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### Llama 4 Maverick (Llama-4-Maverick...) Average FCR: 9.5. Result ID: 699624b85698665ec4656c71. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### Gemma 3 27B (gemma-3-27b-it) Average FCR: 10.8. Result ID: 699624b85698665ec4656c74. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### Mistral Small 24B (Mistral-Small-24B...) Average FCR: 12. Result ID: 699624b85698665ec4656c77. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### Gemini Flash Latest (gemini-flash-preview...) Average FCR: 13.2. Result ID: 699624b85698665ec4656c7a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### Claude 4.1 Opus (claude-opus-4-1...) Average FCR: 14.8. Result ID: 699624b85698665ec4656c7d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Canonical RAG"}. This record does not assert independent reproduction. ### GPT-5.2 (gpt-5.2-2025-12-11) Average FCR: 15.1. Result ID: 699624b85698665ec4656c5b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### Gemini 3.0 Pro (gemini-3.0-pro) Average FCR: 16.8. Result ID: 699624b85698665ec4656c5e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### GPT-4.1 (gpt-4.1-2025-04-14) Average FCR: 18.5. Result ID: 699624b85698665ec4656c61. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### Kimi K2 Thinking (Kimi-K2-Thinking) Average FCR: 21.9. Result ID: 699624b85698665ec4656c67. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### DeepSeek V3.1 Terminus (deepseek-v3.1-terminus) Average FCR: 22.4. Result ID: 699624b85698665ec4656c64. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### Qwen3 235B A22B Instruct (Qwen3-235B-A22B...) Average FCR: 26.5. Result ID: 699624b85698665ec4656c6a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### Claude 4.5 Sonnet (claude-sonnet-4-5...) Average FCR: 28.2. Result ID: 699624b85698665ec4656c6d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### Gemma 3 27B (gemma-3-27b-it) Average FCR: 30.2. Result ID: 699624b85698665ec4656c73. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### Llama 4 Maverick (Llama-4-Maverick...) Average FCR: 31. Result ID: 699624b85698665ec4656c70. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### Mistral Small 24B (Mistral-Small-24B...) Average FCR: 34.5. Result ID: 699624b85698665ec4656c76. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### Gemini Flash Latest (gemini-flash-preview...) Average FCR: 36.8. Result ID: 699624b85698665ec4656c79. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. ### Claude 4.1 Opus (claude-opus-4-1...) Average FCR: 40. Result ID: 699624b85698665ec4656c7c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Paradigm":"Direct"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.