# Best AI for Financial Verifiability (November 2026)

The leader on each benchmark in this use case, never combined. Data refreshed 2 Nov 2026.

## Picks

- ClaimBench: d1-ultra: 100 accuracy % (Vals.ai leaderboard, 23 Oct 2026), https://benchmarks.finance/leaderboards/vals-claimbench/
- FAITH: Claude Sonnet 4: 95.6 overall accuracy % (Paper Table 3, 24 Oct 2025), https://benchmarks.finance/leaderboards/faith/
- K-FinHallu: Gemini 2.5 Flash: 0.860 binary detection F1 (Paper Table 3, 28 May 2026), https://benchmarks.finance/leaderboards/k-finhallu/
- FinLFQA: GPT-4o: 56.5 evidence attribution F1 (Paper Table 6, 7 Oct 2025), https://benchmarks.finance/leaderboards/finlfqa/
- Moneda CiteBench: Numa 1: 63.3 Factual share of linked claims % (Moneda run, 2026-11-02, 2 Nov 2026), https://benchmarks.finance/leaderboards/moneda-citebench/

## ClaimBench: accuracy % (higher is better, 10 models)

| Rank | Model | Provider | Score | Metric | Setting | Reported by | Source | URL | Date |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | d1-ultra | Not stated | 100 | accuracy % | decision models board | Official board | Vals.ai leaderboard | https://www.vals.ai/benchmarks/claimbench | 23 Oct 2026 (checked 1 Nov 2026) |
| 2 | Decider V1.2 | Not stated | 99.7 | accuracy % | decision models board | Official board | Vals.ai leaderboard | https://www.vals.ai/benchmarks/claimbench | 23 Oct 2026 (checked 1 Nov 2026) |
| 3 | Drex Max | Not stated | 99.2 | accuracy % | decision models board | Official board | Vals.ai leaderboard | https://www.vals.ai/benchmarks/claimbench | 23 Oct 2026 (checked 1 Nov 2026) |
| 4 | Mercury Decide | Not stated | 99.1 | accuracy % | decision models board | Official board | Vals.ai leaderboard | https://www.vals.ai/benchmarks/claimbench | 23 Oct 2026 (checked 1 Nov 2026) |
| 5 | Decider V1.1 27B | Not stated | 98.8 | accuracy % | decision models board | Official board | Vals.ai leaderboard | https://www.vals.ai/benchmarks/claimbench | 23 Oct 2026 (checked 1 Nov 2026) |
| 6 | d1 | Not stated | 98.7 | accuracy % | decision models board | Official board | Vals.ai leaderboard | https://www.vals.ai/benchmarks/claimbench | 23 Oct 2026 (checked 1 Nov 2026) |
| 7 | Microsoft-Decision-1 | Microsoft | 98.2 | accuracy % | decision models board | Official board | Vals.ai leaderboard | https://www.vals.ai/benchmarks/claimbench | 23 Oct 2026 (checked 1 Nov 2026) |
| 8 | Jev 1.14 | Not stated | 98.1 | accuracy % | decision models board | Official board | Vals.ai leaderboard | https://www.vals.ai/benchmarks/claimbench | 23 Oct 2026 (checked 1 Nov 2026) |
| 9 | Drex v1.5 | Not stated | 98.0 | accuracy % | decision models board | Official board | Vals.ai leaderboard | https://www.vals.ai/benchmarks/claimbench | 23 Oct 2026 (checked 1 Nov 2026) |
| 10 | Jev 1.13.0 | Not stated | 97.7 | accuracy % | decision models board | Official board | Vals.ai leaderboard | https://www.vals.ai/benchmarks/claimbench | 23 Oct 2026 (checked 1 Nov 2026) |

## FAITH: overall accuracy % (higher is better, 10 models)

| Rank | Model | Provider | Score | Metric | Setting | Reported by | Source | URL | Date |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | Claude Sonnet 4 | Anthropic | 95.6 | overall accuracy % | Main split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2508.05201v2 | 24 Oct 2025 (checked 1 Nov 2026) |
| 2 | Gemini 2.5 Pro | Google | 91.9 | overall accuracy % | Main split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2508.05201v2 | 24 Oct 2025 (checked 1 Nov 2026) |
| 3 | GPT-4.1 | OpenAI | 89.2 | overall accuracy % | Main split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2508.05201v2 | 24 Oct 2025 (checked 1 Nov 2026) |
| 4 | Gemini 2.5 Flash | Google | 88.7 | overall accuracy % | Main split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2508.05201v2 | 24 Oct 2025 (checked 1 Nov 2026) |
| 5 | GPT-4.1 mini | OpenAI | 88.2 | overall accuracy % | Main split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2508.05201v2 | 24 Oct 2025 (checked 1 Nov 2026) |
| 6 | Claude-haiku-3 | Anthropic | 81.3 | overall accuracy % | Main split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2508.05201v2 | 24 Oct 2025 (checked 1 Nov 2026) |
| 7 | Qwen3-32B | Qwen | 73.9 | overall accuracy % | Main split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2508.05201v2 | 24 Oct 2025 (checked 1 Nov 2026) |
| 8 | GPT-4.1-nano | OpenAI | 70 | overall accuracy % | Main split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2508.05201v2 | 24 Oct 2025 (checked 1 Nov 2026) |
| 9 | Gemini-2.5-flash-lite | Google | 50.2 | overall accuracy % | Main split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2508.05201v2 | 24 Oct 2025 (checked 1 Nov 2026) |
| 10 | Llama-3.1-8B | Meta | 47.5 | overall accuracy % | Main split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2508.05201v2 | 24 Oct 2025 (checked 1 Nov 2026) |

## K-FinHallu: binary detection F1 (higher is better, 10 models)

| Rank | Model | Provider | Score | Metric | Setting | Reported by | Source | URL | Date |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | Gemini 2.5 Flash | Google | 0.860 | binary detection F1 | test split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2605.29523v1 | 28 May 2026 (checked 1 Nov 2026) |
| 2 | Gemini 2.5 Pro | Google | 0.849 | binary detection F1 | test split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2605.29523v1 | 28 May 2026 (checked 1 Nov 2026) |
| 3 | GPT-5 | OpenAI | 0.823 | binary detection F1 | test split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2605.29523v1 | 28 May 2026 (checked 1 Nov 2026) |
| 4 | Llama-3.3-70B | Meta | 0.756 | binary detection F1 | test split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2605.29523v1 | 28 May 2026 (checked 1 Nov 2026) |
| 5 | Llama-3.1-70B | Meta | 0.735 | binary detection F1 | test split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2605.29523v1 | 28 May 2026 (checked 1 Nov 2026) |
| 6 | GPT-4o | OpenAI | 0.733 | binary detection F1 | test split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2605.29523v1 | 28 May 2026 (checked 1 Nov 2026) |
| 7 | Qwen3-32B | Qwen | 0.732 | binary detection F1 | test split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2605.29523v1 | 28 May 2026 (checked 1 Nov 2026) |
| 8 | EXAONE-4.0-32B | LG | 0.728 | binary detection F1 | test split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2605.29523v1 | 28 May 2026 (checked 1 Nov 2026) |
| 9 | Qwen3-14B | Qwen | 0.707 | binary detection F1 | test split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2605.29523v1 | 28 May 2026 (checked 1 Nov 2026) |
| 10 | Llama-3.1-8B | Meta | 0.674 | binary detection F1 | test split | Reported by the authors | Paper Table 3 | https://arxiv.org/html/2605.29523v1 | 28 May 2026 (checked 1 Nov 2026) |

## FinLFQA: evidence attribution F1 (higher is better, 8 models)

| Rank | Model | Provider | Score | Metric | Setting | Reported by | Source | URL | Date |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | GPT-4o | OpenAI | 56.5 | evidence attribution F1 | post-hoc attribution, test set | Reported by the authors | Paper Table 6 | https://arxiv.org/html/2510.06426v1 | 7 Oct 2025 (checked 1 Nov 2026) |
| 2 | Mistral-Small-24B | Mistral AI | 55.3 | evidence attribution F1 | post-hoc attribution, test set | Reported by the authors | Paper Table 6 | https://arxiv.org/html/2510.06426v1 | 7 Oct 2025 (checked 1 Nov 2026) |
| 3 | Llama-3.3-70B | Meta | 52.9 | evidence attribution F1 | post-hoc attribution, test set | Reported by the authors | Paper Table 6 | https://arxiv.org/html/2510.06426v1 | 7 Oct 2025 (checked 1 Nov 2026) |
| 4 | Qwen2.5-72B | Qwen | 46.6 | evidence attribution F1 | post-hoc attribution, test set | Reported by the authors | Paper Table 6 | https://arxiv.org/html/2510.06426v1 | 7 Oct 2025 (checked 1 Nov 2026) |
| 5 | phi-4 | Microsoft | 46.1 | evidence attribution F1 | post-hoc attribution, test set | Reported by the authors | Paper Table 6 | https://arxiv.org/html/2510.06426v1 | 7 Oct 2025 (checked 1 Nov 2026) |
| 6 | Mistral-8x22B | Mistral AI | 42.9 | evidence attribution F1 | post-hoc attribution, test set | Reported by the authors | Paper Table 6 | https://arxiv.org/html/2510.06426v1 | 7 Oct 2025 (checked 1 Nov 2026) |
| 7 | Llama-3.2-3B | Meta | 11.8 | evidence attribution F1 | post-hoc attribution, test set | Reported by the authors | Paper Table 6 | https://arxiv.org/html/2510.06426v1 | 7 Oct 2025 (checked 1 Nov 2026) |
| 8 | Llama-3.2-1B | Meta | 1.5 | evidence attribution F1 | post-hoc attribution, test set | Reported by the authors | Paper Table 6 | https://arxiv.org/html/2510.06426v1 | 7 Oct 2025 (checked 1 Nov 2026) |

## Moneda CiteBench: Factual share of linked claims % (higher is better, 1 model)

| Rank | Model | Provider | Score | Metric | Setting | Reported by | Source | URL | Date |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | Numa 1 | Moneda | 63.3 | Factual share of linked claims % | Experimental, not yet human-audited: 100 research notes, a weighted sample of claims judged by GPT-5.4 and Claude Sonnet 4.6 against the cited page. A 50-claim model audit found the judges strict (estimate about 80% if corrected). 72.5% of sentences carry no link. | Run by Moneda | Moneda run, 2026-11-02 | https://benchmarks.finance/method/ | 2 Nov 2026 |


Source: https://benchmarks.finance/best/verifiability/

Created and maintained by Moneda.