WebDev Arena ranks AI models on front-end web development tasks and multi-step coding workflows.
Vals AI
Vals AI evaluates models on real-world tasks across fields including finance, software, cybersecurity, and healthcare.
About Vals AI
Vals AI publishes evaluations of AI models on real-world tasks in fields including finance, software, cybersecurity, healthcare, and legal work. The company says it runs its own evaluations and creates many of its benchmarks in-house. Its homepage shows a model leaderboard with performance scores, cost, latency, and token usage, along with recent reports on individual model releases.
Reports cover named benchmarks such as the Vals Index, CyberBench, LegalBench, Vibe Code Bench, and SAFE-Teen. SAFE-Teen tests how AI chatbots respond to teens seeking mental-health support, using 101 clinician-written scenarios with simulated teens aged 13 to 17. Results are tied to the specific model configurations and settings used in each run.
Key features
- Model leaderboard with performance score, cost, latency, and token usage columns
- Benchmarks covering finance, software, cybersecurity, healthcare, and legal tasks
- Results on the Vals Index, CyberBench, LegalBench, and Vibe Code Bench
- SAFE-Teen benchmark for chatbot responses to teen mental-health conversations
- Per-test cost and per-million-token prices reported in model evaluation posts
- Comparison of single-agent and multi-agent team performance on coding tasks
Best for: People comparing AI models on specific professional tasks, such as legal, financial, software, or cybersecurity work.
Good for
- States that it runs its own evaluations and creates many benchmarks in-house
- Reports cost, latency, and token usage alongside performance scores
- Publishes results for named benchmarks with model rankings and scores
Watch out for
- Some reported runs use specific settings, such as reasoning effort set to max
- Ranking counts differ between benchmarks, so results are not directly comparable across all of them
- The Haiku 5.5 run had no refusal fallback, so content-filter refusals were scored as failures
Vals AI at a glance
- Website
- vals.ai
- Category
- AI & Machine Learning › AI Data & Analytics
- Type
- Website
- Pricing
- Not confirmed. Check the site
- Platforms
- Web
- Open source
- Not stated
- Language
- English
- Last checked
- Oct 6, 2026 · Online when checked
- Listed
- Oct 5, 2026
Vals AI alternatives
Tools and sites you could use instead, ranked by how closely they match.
OpenRouter publishes language model rankings based on real usage through its API.
SandboxAQ applies artificial intelligence and computing to challenges including quantitative models and scientific research.
Compare and rank text-to-image models on the Text-to-Image Arena leaderboard.
Compare outputs from image generation models and view their relative rankings in this arena.
LLM Stats compares language models by rankings for intelligence, speed, and price.
Popular in AI Data & Analytics
Well-known listings from the same topic.
Dotgetdot.aiForvioforvio.comFluidAI Medicalfluidai.mdApache Zeppelinzeppelin.apache.orgSola Securitysola.securityQuadraticquadratichq.com
Vals AI FAQ
What does the SAFE-Teen benchmark test?
It tests how AI chatbots respond when a teen turns to them for mental-health support. The benchmark uses 101 clinician-written scenarios with simulated teens aged 13 to 17, and rates each conversation on clinician-defined safety checks.
Who runs the evaluations on Vals AI?
The site states that Vals AI runs all of its own evaluations and creates many of its benchmarks in-house.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.