Benchmark profile

Vals CorpFin v2 (CorpFin v2)

Vals AI private benchmark for understanding long-context credit agreements.

Data verified July 22, 2026

How BenchLM shows CorpFin v2

BenchLM mirrors the public Vals AI CorpFin v2 leaderboard captured from https://www.vals.ai/benchmarks/corp_fin_v2 and updated by Vals on July 22, 2026. The snapshot preserves overall scores, uncertainty, latency, cost-per-test metadata, and task-level scores where Vals publishes them.

CorpFin v2 is display only on BenchLM. Vals proprietary or Vals-hosted aggregate views are useful context, but BenchLM does not use them as weighted ranking inputs or as a replacement for benchmark-native source records.

126 Vals rows4 task viewsprivate datasetTasks: Overall, Exact Pages, Max Fitting Context, Shared Max ContextDisplay only

CorpFin v2 on Vals AI Vals methodology Vals home

CorpFin v2 score on CorpFin v2 — July 22, 2026

BenchLM mirrors the published corpfin v2 score view for CorpFin v2. Claude Opus 5 leads the public snapshot at 73.19% , followed by Claude Fable 5 (71.83%) and Kimi K3 (71.56%). BenchLM does not use these results to rank models overall.

Claude Opus 5

Anthropic

anthropic/claude-opus-5

73.19%

Overall —

Claude Fable 5

Anthropic

anthropic/claude-fable-5

71.83%

Overall —

Kimi K3

Moonshot AI

kimi/kimi-k3

71.56%

Overall —

126 modelsExternal benchmark mirrorsCurrentDisplay onlyUpdated July 22, 2026

CorpFin v2 score table (126 models)

Score

Claude Opus 5Anthropic

73.19%

Claude Fable 5Anthropic

71.83%

Kimi K3Moonshot AI

71.56%

Muse Spark 1.1Meta

71.29%

InklingThinkingmachines

68.57%

Grok 4.3xAI

68.53%

GPT-5.5OpenAI

68.42%

Kimi K2.5 ThinkingMoonshot AI

68.26%

MiniMax M3MiniMax

68.10%

Qwen3 MaxAlibaba

68.03%

Grok 4.5xAI

67.41%

Claude Opus 4.6 ThinkingAnthropic

67.02%

Claude Sonnet 5Anthropic

66.98%

Grok 4 Fast ReasoningxAI

66.90%

Kimi K2.6Moonshot AI

66.74%

Claude Opus 4.8Anthropic

66.71%

Qwen3.6 Max PreviewAlibaba

66.47%

Gemini 3 Flash PreviewGoogle

66.43%

GLM 5.2Zhipu AI

66.12%

Claude Opus 4.7Anthropic

66.08%

Grok 4 0709xAI

66.05%

Grok 4.1 Fast ReasoningxAI

65.97%

GPT-5.2OpenAI

65.89%

Nemotron 3 Ultra 550b A55bNvidia

65.46%

GPT-5.6 TerraOpenAI

65.31%

Claude Sonnet 4.6Anthropic

65.31%

Qwen3.5 Plus ThinkingAlibaba

65.31%

GPT-5.4OpenAI

65.27%

Muse SparkMeta

65.11%

Claude Opus 4.5 20251101 ThinkingAnthropic

65.07%

Gemini 3.5 FlashGoogle

64.69%

Gemini 3.1 Pro PreviewGoogle

64.49%

GLM 5.1Zhipu AI

64.45%

GPT-5.6 SolOpenAI

64.38%

GPT-5.6 LunaOpenAI

64.22%

GPT-5.1OpenAI

63.83%

Qwen3.7 MaxAlibaba

63.71%

Grok 4.20 0309 ReasoningxAI

63.67%

Gemini 3 Pro PreviewGoogle

63.67%

Qwen3.5 FlashAlibaba

63.56%

Gemini 3.6 FlashGoogle

63.33%

GPT-4.1OpenAI

63.05%

GLM 5 ThinkingZhipu AI

62.90%

Qwen3.6 27bAlibaba

62.31%

Claude Sonnet 4.5 20250929 ThinkingAnthropic

61.97%

Qwen3.6 PlusAlibaba

61.93%

Mimo V2.5 ProXiaomi

61.42%

DeepSeek V4 ProDeepSeek

61.38%

Claude Opus 4.5Anthropic

61.30%

Claude Sonnet 4 20250514 ThinkingAnthropic

61.23%

GPT-5.4 NanoOpenAI

61.19%

MiniMax M2.7MiniMax

61.19%

Grok 3 Mini Fast High ReasoningxAI

61.11%

GPT-5OpenAI

61.07%

Mistral Large 2512Mistral AI

61.03%

GLM 4.5Zhipu AI

60.96%

GPT-5.4 MiniOpenAI

60.92%

Claude Sonnet 4.5Anthropic

60.80%

Gemini 2.5 ProGoogle

60.80%

Gemini 3.5 Flash LiteGoogle

60.68%

Claude Haiku 4.5 20251001 ThinkingAnthropic

60.61%

Kimi K2 ThinkingMoonshot AI

60.57%

Claude 3.7 Sonnet 20250219 ThinkingAnthropic

60.41%

Claude Haiku 4.5Anthropic

60.30%

GPT-5 MiniOpenAI

60.18%

Mimo V2.5Xiaomi

59.91%

Gemini 2.5 Pro Exp 03 25Google

59.83%

Gemini 2.5 Flash Preview 09 2025 ThinkingGoogle

59.75%

O3OpenAI

59.71%

Grok 3xAI

59.71%

MiniMax M2.5MiniMax

59.60%

Grok 3 Mini Fast Low ReasoningxAI

59.48%

Gemini 3.1 Flash Lite PreviewGoogle

59.36%

O4 MiniOpenAI

58.97%

Gemini 2.5 Flash Preview 09 2025Google

58.97%

MiniMax M2.1MiniMax

58.90%

Mistral Medium 3.5Mistral AI

58.78%

Grok 4 Fast Non ReasoningxAI

58.39%

GPT Oss 120bFireworks AI

58.24%

Laguna M.1Poolside

58.16%

GPT-4.1 MiniOpenAI

57.93%

Gemini 2.5 Flash Lite Preview 09 2025 ThinkingGoogle

57.58%

GLM 4.6Zhipu AI

56.84%

Laguna Xs.2Poolside

56.33%

Gemini 2.5 Flash Lite Preview 09 2025Google

56.29%

Qwen3 MaxAlibaba

55.94%

DeepSeek V3 0324Fireworks AI

54.74%

Claude Sonnet 4Anthropic

54.70%

Trinity Large ThinkingArcee-Ai

54.66%

Gemini 2.5 Flash Preview 04 17Google

54.16%

DeepSeek R1Fireworks AI

54.12%

Claude 3.5 SonnetAnthropic

53.61%

GPT Oss 20bFireworks AI

53.15%

Qwen3 Max PreviewAlibaba

52.95%

Grok 4.1 Fast Non ReasoningxAI

52.49%

DeepSeek V3Fireworks AI

52.49%

DeepSeek V3p1Fireworks AI

51.48%

Grok 2 1212xAI

51.13%

DeepSeek V3p2 ThinkingFireworks AI

50.97%

100

Claude 3.5 HaikuAnthropic

50.82%

101

Mistral Medium 2505Mistral AI

50.74%

102

Moonshotai Kimi K2 InstructTogether AI

50.39%

103

Llama4 Maverick Instruct BasicFireworks AI

49.73%

104

DeepSeek V3p2Fireworks AI

47.94%

105

Magistral Medium 2509Mistral AI

47.40%

106

Meta Llama Llama 4 Scout 17B 16E InstructTogether AI

46.78%

107

GLM 4.7Zhipu AI

46.39%

108

Command A 03 2025Cohere

45.96%

109

GPT-4oOpenAI

45.92%

110

GPT-4o MiniOpenAI

45.45%

111

O3 MiniOpenAI

45.30%

112

Mistral Small 2503Mistral AI

44.17%

113

Magistral Small 2509Mistral AI

44.02%

114

Gemini 2.0 Pro Exp 02 05Google

43.44%

115

GPT-4.1 NanoOpenAI

42.08%

116

Jamba Large 1.6AI21 Labs

41.53%

117

Gemini 1.5 Pro 002Google

40.52%

118

GPT-4oOpenAI

39.43%

119

Jamba 1.5 LargeAI21 Labs

39.43%

120

Meta Llama Meta Llama 3.1 70B Instruct TurboTogether AI

38.85%

121

Gemini 1.5 Flash 002Google

38.19%

122

Jamba Mini 1.6AI21 Labs

38.03%

123

Meta Llama Meta Llama 3.1 8B Instruct TurboTogether AI

37.80%

124

Jamba 1.5 MiniAI21 Labs

33.88%

125

Gemini 2.0 Flash 001Google

33.72%

126

Gemini 1.5 Flash 001Google

28.63%

The published CorpFin v2 snapshot places Claude Opus 5 first at 73.19%. The third row is 1.63 points behind. The broader top-10 range is 5.17 points, so many of the published results sit in a relatively narrow band.

126 models have been evaluated on CorpFin v2. The benchmark falls in the External benchmark mirrors category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. CorpFin v2 is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About CorpFin v2

Year

2026

Tasks

Credit-agreement understanding tasks

Format

Accuracy score

Difficulty

Professional finance document reasoning

The Vals CorpFin v2 page reports overall, exact-page, max-fitting-context, and shared-max-context task views. BenchLM keeps it display only.

CorpFin v2 Public benchmark source

BenchLM freshness & provenance

Version

CorpFin v2 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

FAQ

What does CorpFin v2 measure?

Vals AI private benchmark for understanding long-context credit agreements.

Which model leads the published CorpFin v2 snapshot?

Claude Opus 5 currently leads the published CorpFin v2 snapshot with 73.19% corpfin v2 score. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on CorpFin v2?

126 AI models are included in BenchLM's mirrored CorpFin v2 snapshot, based on the public leaderboard captured on July 22, 2026.

Last updated: July 22, 2026 · mirrored from the public benchmark leaderboard

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 2,000+ readers.

One email each week. Unsubscribe anytime.