Vals Index v2 (Vals Index)
We show this table for reference; we do not rank on it.
Vals AI composite benchmark across professional finance, coding, modeling, and legal-work tasks, including Finance Agent v2, EMB, Terminal-Bench 2.1, Vibe Code Bench, Code Migration, Legal Research, and HLAB.
Evaluation results published by Vals AI. Published leaderboard aggregates and evaluation settings only. Dataset contents are not included.
Vals Index score on Vals Index — October 6, 2026
We mirror the published vals index score view for Vals Index. Gemini 4 Argon leads the public snapshot at 68.90%, followed by Claude Sonnet 5.5 (67.04%) and Claude Opus 5.5 (66.97%). We do not use these results to rank models overall.
Gemini 4 Argon
high reasoning
Claude Sonnet 5.5
Anthropic
Claude Opus 5.5
Anthropic
44 modelsCross-domainCurrentDisplay onlyUpdated October 6, 2026
Vals Index score table (44 models)
| Rank | Model / configuration | Score | Parameters (B) | Open / closed |
|---|---|---|---|---|
| 1 | Gemini 4 ArgonGooglehigh reasoning | 68.90% | Not reported | Closed |
| 2 | Claude Sonnet 5.5Anthropic | 67.04% | Not reported | Closed |
| 3 | Claude Opus 5.5Anthropic | 66.97% | Not reported | Closed |
| 4 | Claude Fable 5.1Anthropic | 65.83% | Not reported | Closed |
| 5 | Claude Opus 5Anthropic | 63.67% | Not reported | Closed |
| 6 | GPT-6 AstraOpenAImax reasoning | 63.13% | Not reported | Closed |
| 7 | Claude Fable 5Anthropic | 61.39% | Not reported | Closed |
| 8 | GPT-6.1 SolOpenAImax reasoning | 61.15% | Not reported | Closed |
| 9 | Muse Spark 1.3 MaxMetamax reasoning | 58.16% | Not reported | Unknown |
| 10 | GPT-5.6 SolOpenAImax reasoning | 58.01% | Not reported | Closed |
| 11 | GPT-6 SolOpenAImax reasoning | 57.54% | Not reported | Closed |
| 12 | MiMo-V2.6-ProXiaomi | 55.20% | Not reported | Open |
| 13 | Claude Opus 4.8Anthropic | 55.10% | Not reported | Closed |
| 14 | Grok 4.7xAIxhigh reasoning | 54.95% | Not reported | Closed |
| 15 | Gemini 3.8 FlashGooglehigh reasoning | 54.83% | Not reported | Closed |
| 16 | GLM-5.3Z.AImax reasoning | 53.51% | Not reported | Open |
| 17 | MiMo-V2.6-FlashXiaomi | 53.23% | Not reported | Open |
| 18 | Muse Spark 1.3Metaxhigh reasoning | 53.20% | Not reported | Closed |
| 19 | GPT-5.6 TerraOpenAImax reasoning | 53.09% | Not reported | Closed |
| 20 | Grok 4.6xAIhigh reasoning | 52.09% | Not reported | Closed |
| 21 | Claude Sonnet 5Anthropic | 51.77% | Not reported | Closed |
| 22 | GPT-5.6 LunaOpenAImax reasoning | 51.69% | Not reported | Closed |
| 23 | DeepSeek V4.1 FlashDeepSeekhigh reasoning | 51.32% | Not reported | Open |
| 24 | Gemini 3.7 FlashGooglehigh reasoning | 51.27% | Not reported | Closed |
| 25 | GPT-6 LunaOpenAImax reasoning | 51.22% | Not reported | Closed |
| 26 | Ember-1Fireworks | 50.81% | Not reported | Closed |
| 27 | Kimi K3Moonshot AImax reasoning | 50.30% | Not reported | Pending |
| 28 | Hy4 previewTencent | 49.94% | Not reported | Open |
| 29 | Step 5 PreviewStepFun | 49.35% | Not reported | Pending |
| 30 | Muse Spark 1.2Metaxhigh reasoning | 49.29% | Not reported | Closed |
| 31 | Qwen3.8 MaxAlibaba | 48.27% | Not reported | Open |
| 32 | Mistral Large 4Mistralhigh reasoning | 48.05% | Not reported | Pending |
| 33 | DeepSeek V4 Flash 0731DeepSeekhigh reasoning | 47.96% | Not reported | Open |
| 34 | DeepSeek V4 Pro 0813DeepSeekmax reasoning | 47.63% | Not reported | Open |
| 35 | Gemini 3.5 FlashGooglehigh reasoning | 44.79% | Not reported | Closed |
| 36 | Grok 4.5xAIhigh reasoning | 44.72% | Not reported | Closed |
| 37 | DeepSeek V4 Pro 0813DeepSeekmax reasoning | 38.63% | Not reported | Open |
| 38 | MiniMax M3MiniMax | 36.53% | Not reported | Open |
| 39 | MiMo-V2.5-ProXiaomi | 33.84% | Not reported | Closed |
| 40 | Gemini 3.1 Pro PreviewGooglehigh reasoning | 33.44% | Not reported | Unknown |
| 41 | GPT-5.4 miniOpenAIxhigh reasoning | 33.17% | Not reported | Closed |
| 42 | InklingThinking Machines Lab0.99 reasoning | 28.67% | Not reported | Open |
| 43 | Inkling-SmallThinking Machines Lab0.99 reasoning | 25.46% | Not reported | Open |
| 44 | Mercury 2.5Inceptionhigh reasoning | 8.86% | Not reported | Closed |
How Vals Index is shown here
BenchLM mirrors the public Vals AI Vals Index leaderboard captured from https://www.vals.ai/benchmarks/vals_index and updated by Vals on October 6, 2026. The snapshot preserves overall scores, uncertainty, latency, cost-per-test metadata, and task-level scores where Vals publishes them.
Vals Index 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
The published Vals Index snapshot places Gemini 4 Argon first at 68.90%. The third row is 1.92 points behind. The broader top-10 range is 10.89 points, so the table still separates the published systems.
44 models have been evaluated on Vals Index. The benchmark falls in the Cross-domain category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. Vals Index is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About Vals Index
Year
2026
Tasks
Finance, coding, spreadsheet modeling, code migration, and legal-work components
Format
Composite score
Difficulty
Private economic-work benchmark composite
We mirror Vals Index v2 as a display-only external composite. The task mix changed from the earlier index, so the current score should not be treated as directly interchangeable with an older snapshot. Vals proprietary aggregate scores remain outside weighted rankings.
Freshness and provenance
Version
Vals Index 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
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 Vals Index measure?
Vals AI composite benchmark across professional finance, coding, modeling, and legal-work tasks, including Finance Agent v2, EMB, Terminal-Bench 2.1, Vibe Code Bench, Code Migration, Legal Research, and HLAB.
Which model leads the published Vals Index snapshot?
Gemini 4 Argon currently leads the published Vals Index snapshot with 68.90% vals index score. BenchLM shows this benchmark for display only and does not use it in overall rankings.
How many models are evaluated on Vals Index?
The October 6, 2026 snapshot contains 44 AI models.
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 5,500+ readers.
One email each week. Unsubscribe anytime.