Professional-task evaluation platform
Vals AI
A view of AI performance grounded in particular kinds of professional work.
Key information
- Focus
- Professional tasks, including finance, software, legal and tax work.
- Measure
- Benchmark-specific task success; Vals Index is a separate weighted aggregate.
- Version / scope
- Vals Index v2.1; other benchmarks versioned separately
- Update pattern
- Per-benchmark releases and model evaluations; no common fixed interval was confirmed.
- Publisher update
- 13 October 2026
- Editorial verification
- 15 October 2026
In plain English
What makes it different?
Vals AI evaluates models on work-oriented tasks and publishes individual benchmarks alongside an aggregate Vals Index. The index weights sectors using a stated economic proxy. It is one interpretation of usefulness, not a forecast of what a model will deliver in every workplace.
Sources and attribution are listed below.
What it measures
Professional tasks, including finance, software, legal and tax work.
Strengths
- Connects evaluation to specific professional workflows.
- Publishes per-benchmark methods rather than only an aggregate.
Limitations
- Several datasets are private, limiting independent reproduction.
- Sector weights and selected tasks do not represent every job, country or organisation. Some published runs use fallback models, which must be examined.
How the evaluation works
Evaluation depends on the selected benchmark: tasks, tools, scoring and permitted assistance need to be checked separately.
For the reviewed Vals Index, component benchmark scores are combined by sector. The methodology explains its weights and version changes; those choices influence the result.
Configuration matters
Compare the same benchmark version and agent setup. Check tools, reasoning settings, refusal handling and any fallback model; a combined workflow is not evidence about the named model alone.
Primary methods: Vals Index methodology and version history ↗
Results at the publisher
Scores stay with the original source
We have not documented the permissions needed to reproduce this leaderboard here. You can still explore its methods and visit the publisher for results.
Versions & freshness
Verified 15 October 2026The reviewed Vals Index page labels version 2.1 and an update on 13 October 2026. That date belongs to the publisher's index page, not to every underlying task run.
Official site, index methodology and update notes accessible. Per-benchmark releases and model evaluations; no common fixed interval was confirmed.
Our review date is not an evaluation date. Individual run dates are not consistently available; they remain unknown here. No automatic result collection is enabled.
Publication & reuse
The reviewed index is marked proprietary. No applicable data-reproduction grant or agreement is documented; its website visibility is not permission to copy scores, task material or history.
Licences of component benchmarks do not automatically license Vals AI's runs or aggregate.
View the scope of the reuse decision
- Result fields
- No result fields cleared
- Public result display
- not enabled
- Storage
- not enabled
- Historical copies
- not enabled
- Updates
- not enabled
- Commercial reuse
- not enabled
- Derived ratings
- No scoped grant documented; not enabled.
- Attribution
- Credit the named publisher and link the primary sources. Attribution alone does not grant reproduction rights.
This decision does not clear mixed-source scientific recommendations. Publisher content, software, datasets and model licences are separate. No partnership is implied.
Primary sources & attribution
This guide is original scor.es commentary about Vals AI's work. The links below support the identity, methods, versions and reuse review.
- Official website ↗Checked 15 October 2026
- Vals Index methodology and version history ↗Checked 15 October 2026
- About Vals AI ↗Checked 15 October 2026
Editorial edition 2026-10-15.1 · reuse review directory-rights-1. Re-check due 14 November 2026.