Skip to main content
Radar

Every change to the models you run, with its source and its date. Releases, price changes, retirements, API changes, and incidents.Every change to the models you run, with its source.

Follow model changes
BenchLM

Finance Agent v2

We mirror this table; we do not rank on it.

Data verified 41 confirmed releases in the last 30 daysFollow model changes

Vals AI benchmark for realistic financial analyst agent tasks across qualitative analysis, quantitative analysis, market work, comparables, precedents, earnings, disclosure, and modeling.

Finance Agent v2 score on Finance Agent v2 — September 21, 2026

We mirror the published finance agent v2 score view for Finance Agent v2. Gemini 3.8 Flash leads the public snapshot at 61.4%, followed by Muse Spark 1.2 (60.6%) and Muse Spark 1.3 Max (60.0%). We do not use these results to rank models overall.

1Proprietary

Gemini 3.8 Flash

Google

high reasoning

61.4%
Overall 74.05Context 1M
2Proprietary

Muse Spark 1.2

Meta

xhigh reasoning

60.6%
Overall 70.01Context 1M
3

Muse Spark 1.3 Max

Meta

max reasoning

60.0%

63 modelsAgenticCurrentDisplay onlyUpdated September 21, 2026

Finance Agent v2 score table (63 models)

Score
1
Gemini 3.8 FlashGoogle · Closedhigh reasoning
61.4%
2
Muse Spark 1.2Meta · Closedxhigh reasoning
60.6%
3
Muse Spark 1.3 MaxMetamax reasoning
60.0%
4
Gemini 3.7 FlashGoogle · Closedhigh reasoning
59.0%
5
Muse Spark 1.3Meta · Closedxhigh reasoning
58.9%
6
Claude Fable 5.1Anthropic · Closed
58.9%
7
Claude Opus 5Anthropic · Closed
58.6%
8
Claude Opus 5.5Anthropic · Closed
58.6%
9
Gemini 3.5 FlashGoogle · Closedhigh reasoning
57.9%
10
GLM-5.3-FlashZ.AI · Open weightmax reasoning
57.9%
11
Muse Spark 1.1Meta · Closedxhigh reasoning
57.2%
12
Claude Fable 5Anthropic · Closed
56.3%
13
Gemini 3.6 FlashGoogle · Closedhigh reasoning
56.3%
14
GLM-5.3Z.AI · Open weightmax reasoning
55.8%
15
Hy4 previewTencent · Open weight
55.1%
16
GPT-5.6 LunaOpenAI · Closedmax reasoning
55.0%
17
GPT-5.6 TerraOpenAI · Closedmax reasoning
54.4%
18
Kimi K3Moonshot AI · Closed
54.4%
19
Claude Opus 4.8Anthropic · Closed
53.9%
20
Claude Sonnet 5Anthropic · Closed
53.9%
21
GPT-5.6 SolOpenAI · Closedmax reasoning
53.8%
22
Grok 4.6xAI · Closedhigh reasoning
53.7%
23
GPT-6 AstraOpenAI · Closedmax reasoning
53.5%
24
DeepSeek V4.1 FlashDeepSeek · Open weighthigh reasoning
53.5%
25
Grok 4.7xAI · Closedxhigh reasoning
52.3%
26
GPT-5.5OpenAI · Closedxhigh reasoning
51.8%
27
Claude Opus 4.7Anthropic · Closed
51.5%
28
Claude Sonnet 4.6Anthropic · Closed
51.0%
29
Qwen3.8 MaxAlibaba · Open weight
50.6%
30
DeepSeek V4 Pro 0813DeepSeek · Closedmax reasoning
50.4%
31
GLM-5.2Z.AI · Open weight
49.7%
32
DeepSeek V4 Flash 0731DeepSeek · Closedhigh reasoning
49.5%
33
Qwen3.8-27BAlibaba · Open weightxhigh reasoning
48.6%
34
Grok 4.5xAI · Closedhigh reasoning
48.3%
35
MiniMax M3MiniMax · Open weight
48.3%
36
Qwen3.7 MaxAlibaba · Closed
47.8%
37
Gemini 3.5 Flash-LiteGoogle · Closedhigh reasoning
47.4%
38
InklingThinking Machines Lab · Open weight0.99 reasoning
46.6%
39
GPT-5.4 miniOpenAI · Closedxhigh reasoning
45.4%
40
Kimi K2.6Moonshot AI · Open weight
44.9%
41
GLM-5.1Z.AI · Open weight
44.8%
42
DeepSeek V4 Pro 0813DeepSeek · Closedmax reasoning
44.1%
43
Gemini 3.1 Pro PreviewGooglehigh reasoning
43.0%
44
Gemini 3 Flash PreviewGooglehigh reasoning
42.6%
45
MiMo-V2.5-ProXiaomi · Closed
41.5%
46
Inkling-SmallThinking Machines Lab · Open weight0.99 reasoning
41.3%
47
Qwen3.6 PlusAlibaba · Closed
40.8%
48
Qwen3.7 PlusAlibaba · Closed
38.2%
49
GPT-5.4 nanoOpenAI · Closedhigh reasoning
38.2%
50
Grok 4.3xAI · Closedhigh reasoning
37.7%
51
Nemotron 3 Ultra 550b A55bNvidia
37.7%
52
MiMo-V2.5Xiaomi · Closed
36.7%
53
Kimi K2.5 ThinkingMoonshot AI
35.8%
54
Mistral Medium 3.5Mistral AIhigh reasoning
32.1%
55
Claude Haiku 4.5 ThinkingAnthropic · Closed
31.0%
56
Ling 3.0 Flash 2607Ant
30.3%
57
Gemini 3.1 Flash Lite PreviewGooglehigh reasoning
30.0%
58
Grok 4.20 0309 ReasoningxAI
28.5%
59
MiniMax M2.7MiniMax · Open weight
27.9%
60
Laguna M.1Poolside · Closed
25.0%
61
Mercury 2.5Inception · Closedhigh reasoning
18.6%
62
Nemotron Lightning 3p5 30b A3bFireworks AI
18.5%
63
Laguna XS.2Poolside · Open weight
15.6%

How Finance Agent v2 is shown here

BenchLM mirrors the public Vals AI Finance Agent v2 leaderboard captured from https://www.vals.ai/benchmarks/fabv2 and updated by Vals on September 21, 2026. The snapshot preserves overall scores, uncertainty, latency, cost-per-test metadata, and task-level scores where Vals publishes them.

Finance Agent 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.

Snapshot

63 Vals rows11 task viewsprivate datasetTasks: Overall, All-Pass, General Qualitative Analysis, General Quantitative Analysis, Market AnalysisDisplay only
Finance Agent v2 on Vals AIVals methodologyVals home

The published Finance Agent v2 snapshot places Gemini 3.8 Flash first at 61.4%. The third row is 1.5 points behind. The broader top-10 range is 3.6 points, so many of the published results sit in a relatively narrow band.

63 models have been evaluated on Finance Agent v2. The benchmark falls in the Agentic category. This category carries a 22% weight in BenchLM.ai's overall scoring system. Finance Agent v2 is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About Finance Agent v2

Year

2026

Tasks

Financial analyst task categories

Format

Mean score across repeated runs

Difficulty

Professional expert-task agent workflow

Vals reports Finance Agent v2 as a multi-category benchmark with severity-weighted partial credit and repeated runs per model. BenchLM mirrors the public Vals leaderboard as a display-only expert-task benchmark.

Finance Agent v2Public benchmark source

Freshness and provenance

Version

Finance Agent 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.

Questions

What does Finance Agent v2 measure?

Vals AI benchmark for realistic financial analyst agent tasks across qualitative analysis, quantitative analysis, market work, comparables, precedents, earnings, disclosure, and modeling.

Which model leads the published Finance Agent v2 snapshot?

Gemini 3.8 Flash currently leads the published Finance Agent v2 snapshot with 61.4% finance agent v2 score. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on Finance Agent v2?

The September 21, 2026 snapshot contains 63 AI models.

Last updated: September 21, 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.