Is there any API or site that provides a unified leaderboard of AI models (benchmarks, pricing, metadata, providers)?

[abderbejaoui]

Hi everyone 👋,

I’m searching for a centralized, up-to-date resource (API or website) that aggregates information about AI models from different providers, including:

What I'm looking for:

• A live leaderboard of AI/LLM models
• Coverage of as many benchmarks as possible, such as:
  • MMLU
  • MATH
  • GPQA
  • BBH
  • IFEval
  • MUSR
  • AGI Eval
• Support for multiple providers (OpenAI, Anthropic, Google, Meta, Mistral, xAI, Alibaba, Amazon Bedrock, etc.)
• Additional metadata, such as:
  • Model provider
  • Pricing
  • Context length
  • Model family
  • Release date
  • Capabilities (vision, audio, tools, etc.)
  • API endpoints
  • Versioning
• Ideally accessible through a public API so that I can integrate it into an app.

What I tried:
I’ve already checked:

• Hugging Face Open LLM Leaderboard (+ API)
• LMSYS / Chatbot Arena
• llm-stats
• Provider dashboards
• Amazon Bedrock / Vertex AI / OpenAI model lists

…but none of these provide a unified, comprehensive, live dataset covering both:

• many benchmarks AND
• many providers,
  with extra metadata like pricing.

My question:
👉 Does such a unified leaderboard/API already exist?
If yes, I would really appreciate a link or recommendation.

If not, I’m also open to contributing to a community-driven dataset or project.

Thanks in advance!

[AKash-A007]

I dont think anything like that exist .. you might have to DIY all the things acc to your requirements
i used hugging face leaderboard for this once @abderbejaoui

[akhshsh557-hue]

yeah @AKash-A007 that would be the only way ig

[barakplasma]

https://api.zeroeval.com/leaderboard/models/full?justCanonicals=true

[L0nE-F0x]

Does this get updated frequently?

[KrishPatel1905]

At the moment, there is no single, unified, live API or leaderboard that aggregates AI models across multiple providers with comprehensive benchmark coverage and full metadata (pricing, context length, versioning, capabilities). Existing platforms cover parts of the problem, but a truly centralized, up-to-date solution does not yet exist—representing a clear opportunity for a community-driven initiative.

[UnknownHawkins]

There is currently no unified, live leaderboard or public API that aggregates multiple benchmarks + all major providers (open and closed) + structured metadata like pricing, context length, capabilities, and versioning in one place.

Every existing option (HF Leaderboard, Chatbot Arena, HELM, provider dashboards) solves only a slice of the problem. Anyone building against this today has to manually normalize data across multiple sources.

This is a clear gap in the ecosystem, not a documentation issue. A standardized, community-driven dataset/API would be genuinely valuable, and I’d be interested in contributing if something like this moves forward.