---
title: "FinanceBenchmark: Financial AI Leaderboard 2026"
description: "Independent evaluation and composite rankings of frontier AI models on financial reasoning and forensic accounting tasks."
canonical: "https://financebenchmark.ai/"
updated: "2026-10-20"
type: "text/markdown"
---
# FinanceBenchmark: Financial AI Leaderboard 2026

> Independent composite rankings of leading Artificial Intelligence models on complex financial reasoning and forensic accounting benchmarks.

## Table of Contents
- [1. Top Ranked Financial AI Models](#top-ranked-models)
- [2. Evaluation Methodology Summary](#methodology-summary)
- [3. Benchmark Domains & Categories](#benchmark-domains)
- [4. Platform Statistics](#platform-statistics)
- [5. Resource Links](#resource-links)

## 1. Top Ranked Financial AI Models {#top-ranked-models}

| Rank | Model | Organization | Composite FB Score | Benchmarks Evaluated | License | Context Window |
|:----:|:------|:-------------|:------------------:|:--------------------:|:--------|:--------------:|
| 1 | [Avae 2.0](https://financebenchmark.ai/models/avae-2-0.md) | Avae | **97.0** * | 2/42 | Proprietary | N/A |
| 2 | [Claude Opus 4.7](https://financebenchmark.ai/models/claude-opus-4-7.md) | Anthropic | **73.8** | 10/42 | proprietary | 200K |
| 3 | [GPT-5.5](https://financebenchmark.ai/models/gpt-5-5.md) | OpenAI | **72.2** | 12/42 | proprietary | 128K |
| 4 | [GPT-6 Sol](https://financebenchmark.ai/models/gpt-6-sol.md) | OpenAI | **71.1** * | 1/42 | proprietary | 1M |
| 5 | [DeepSeek V4 Pro](https://financebenchmark.ai/models/deepseek-v4-pro.md) | DeepSeek | **70.1** | 9/42 | open-weight | 128K |
| 6 | [Gemini 3.1 Pro](https://financebenchmark.ai/models/gemini-3-1-pro.md) | Google | **70.0** | 9/42 | proprietary | 1M |
| 7 | [GPT-6 Luna](https://financebenchmark.ai/models/gpt-6-luna.md) | OpenAI | **70.0** * | 1/42 | proprietary | 1M |
| 8 | [Claude Fable 5](https://financebenchmark.ai/models/claude-fable-5.md) | Anthropic | **69.8** | 25/42 | proprietary | 1M |
| 9 | [Kimi K2.6](https://financebenchmark.ai/models/kimi-k2-6.md) | Moonshot AI | **67.4** | 9/42 | open-weight | 128K |
| 10 | [Claude Opus 5](https://financebenchmark.ai/models/claude-opus-5.md) | Anthropic | **66.7** | 30/42 | proprietary | 1M |
| 11 | [Claude Fable 5.1](https://financebenchmark.ai/models/claude-fable-5-1.md) | Anthropic | **66.2** | 5/42 | proprietary | 1M |
| 12 | [GLM-5.1](https://financebenchmark.ai/models/glm-5-1.md) | Zhipu AI | **65.9** | 7/42 | open-weight | 128K |
| 13 | [Grok 4.6](https://financebenchmark.ai/models/grok-4-6.md) | xAI | **65.4** | 21/42 | proprietary | 500K |
| 14 | [GPT-5.6 Sol](https://financebenchmark.ai/models/gpt-5-6-sol.md) | OpenAI | **65.1** | 29/42 | proprietary | 1M |
| 15 | [Grok 4.3](https://financebenchmark.ai/models/grok-4-3.md) | xAI | **65.0** | 9/42 | proprietary | 131K |
| 16 | [Gemini 3.8 Flash](https://financebenchmark.ai/models/gemini-3-8-flash.md) | Google | **64.8** | 3/42 | proprietary | 1M |
| 17 | [GPT-5.6 Luna](https://financebenchmark.ai/models/gpt-5-6-luna.md) | OpenAI | **64.6** | 25/42 | proprietary | 1M |
| 18 | [Claude Sonnet 5](https://financebenchmark.ai/models/claude-sonnet-5.md) | Anthropic | **64.1** | 25/42 | proprietary | 1M |
| 19 | [DeepSeek V4 Pro (0813)](https://financebenchmark.ai/models/deepseek-v4-pro-0813.md) | DeepSeek | **63.7** | 16/42 | open-weight | 1M |
| 20 | [Muse Spark 1.2](https://financebenchmark.ai/models/muse-spark-1-2.md) | Meta | **63.7** | 23/42 | proprietary | 1M |
| 21 | [Kimi K3](https://financebenchmark.ai/models/kimi-k3.md) | Moonshot AI | **63.4** | 27/42 | open-weight | 1M |
| 22 | [GLM 5.3 Flash](https://financebenchmark.ai/models/glm-5-3-flash.md) | Zhipu AI | **63.4** | 3/42 | open-weight | 1M |
| 23 | [DeepSeek V4 Flash](https://financebenchmark.ai/models/deepseek-v4-flash.md) | DeepSeek | **63.2** | 19/42 | open-weight | 1M |
| 24 | [GPT-5.6 Terra](https://financebenchmark.ai/models/gpt-5-6-terra.md) | OpenAI | **63.1** | 25/42 | proprietary | 1M |
| 25 | [Gemini 4 Argon](https://financebenchmark.ai/models/gemini-4-argon.md) | Google | **63.1** | 3/42 | proprietary | 1M |
| 26 | [Muse Spark 1.1](https://financebenchmark.ai/models/muse-spark-1-1.md) | Meta | **62.8** | 23/42 | proprietary | 1M |
| 27 | [Qwen 3.8 Max](https://financebenchmark.ai/models/qwen-3-8-max.md) | Alibaba | **62.2** | 25/42 | open-weight | 1M |
| 28 | [Gemini 3.6 Flash](https://financebenchmark.ai/models/gemini-3-6-flash.md) | Google | **61.6** | 22/42 | proprietary | 1M |
| 29 | [Gemini 3.7 Flash](https://financebenchmark.ai/models/gemini-3-7-flash.md) | Google | **61.2** | 3/42 | proprietary | 1M |
| 30 | [Claude Opus 5.5](https://financebenchmark.ai/models/claude-opus-5-5.md) | Anthropic | **59.4** | 3/42 | proprietary | 1M |
| 31 | [GPT-6 Astra](https://financebenchmark.ai/models/gpt-6-astra.md) | OpenAI | **59.3** | 3/42 | proprietary | 1.1M |
| 32 | [MiniMax M2.7](https://financebenchmark.ai/models/minimax-m2-7.md) | MiniMax | **58.7** | 8/42 | open-weight | 1M |
| 33 | [Claude Sonnet 5.5](https://financebenchmark.ai/models/claude-sonnet-5-5.md) | Anthropic | **58.4** | 3/42 | proprietary | 1M |
| 34 | [Muse Spark 1.3 Max](https://financebenchmark.ai/models/muse-spark-1-3-max.md) | Meta | **58.0** | 3/42 | proprietary | 1M |
| 35 | [GLM 5.2](https://financebenchmark.ai/models/glm-5-2.md) | Zhipu AI | **57.5** | 20/42 | open-weight | 1M |
| 36 | [MiniMax M3](https://financebenchmark.ai/models/minimax-m3.md) | MiniMax | **57.2** | 5/42 | open-weight | 1M |
| 37 | [Grok 4.5](https://financebenchmark.ai/models/grok-4-5.md) | xAI | **56.8** | 25/42 | proprietary | 500K |
| 38 | [GLM 5.3](https://financebenchmark.ai/models/glm-5-3.md) | Zhipu AI | **55.3** | 3/42 | open-weight | 1M |
| 39 | [Muse Spark 1.3](https://financebenchmark.ai/models/muse-spark-1-3.md) | Meta | **55.1** | 3/42 | proprietary | 1M |
| 40 | [DeepSeek V4.1 Flash](https://financebenchmark.ai/models/deepseek-v4-1-flash.md) | DeepSeek | **54.8** | 3/42 | open-weight | 1M |
| 41 | [Gemini 3.5 Flash Lite](https://financebenchmark.ai/models/gemini-3-5-flash-lite.md) | Google | **53.5** | 23/42 | proprietary | 1M |
| 42 | [Kimi K2.7 Code](https://financebenchmark.ai/models/kimi-k2-7-code.md) | Moonshot AI | **53.5** | 7/42 | open-weight | 262K |
| 43 | [Inkling](https://financebenchmark.ai/models/inkling.md) | Thinking Machines | **52.3** | 26/42 | open-weight | 256K |
| 44 | [Inkling Small](https://financebenchmark.ai/models/inkling-small.md) | Thinking Machines | **51.5** | 24/42 | open-weight | 262K |
| 45 | [Nemotron 3 Ultra](https://financebenchmark.ai/models/nemotron-3-ultra.md) | NVIDIA | **46.3** | 16/42 | open-weight | 1M |
| 46 | [Ling 3.0 Flash](https://financebenchmark.ai/models/ling-3-0-flash.md) | Ant Group | **45.0** | 18/42 | open-weight | 262K |
| 47 | [Qwen 3.7 Plus](https://financebenchmark.ai/models/qwen-3-7-plus.md) | Alibaba | **39.5** | 12/42 | proprietary | 1M |

*\* Indicates models evaluated on fewer than 3 benchmarks.*

## 2. Evaluation Methodology Summary {#methodology-summary}

$$\text{FB Score} = \frac{\sum (\text{Benchmark Score} \times \text{Tier Weight})}{\sum \text{Tier Weight}}$$

- **Tier 1 (Weight 30):** High-complexity tasks on raw structured financial data, requiring code execution or multi-file forensic verification.
- **Tier 2 (Weight 20):** Numerical reasoning and document comprehension over corporate filings.
- **Tier 3 (Weight 10):** Domain NLP and multi-document audit reasoning.

Read full methodology: [Methodology Document](https://financebenchmark.ai/methodology.md)

## 3. Benchmark Domains & Categories {#benchmark-domains}

- **Verification & Forensic Reasoning:** Tests whether models can verify calculations against ground-truth contracts.
- **Document QA:** Tests factual retrieval and structured synthesis from 10-K, 10-Q, and annual reports.
- **Numerical Reasoning:** Tests multi-step mathematical operations and tabular computations.
- **Agent Tasks:** Evaluates end-to-end autonomous research, code execution, and multi-step tool use.

## 4. Platform Statistics {#platform-statistics}

- **Models Tracked:** 52
- **Benchmarks Included:** 42
- **Total Evaluation Tasks:** 13,500+

## 5. Resource Links {#resource-links}

- [All Models Directory](https://financebenchmark.ai/models.md)
- [All Benchmarks Catalog](https://financebenchmark.ai/benchmarks.md)
- [Research & Analysis](https://financebenchmark.ai/research.md)
- [Submit a Score](https://financebenchmark.ai/submit.md)
- [Curated LLM Index (llms.txt)](https://financebenchmark.ai/llms.txt)
- [Complete Dataset (llms-full.txt)](https://financebenchmark.ai/llms-full.txt)