About ModelSpec
ModelSpec at modelspec.dev is not OpenAI's Model Spec, the CNCF ModelPack model-spec, or the Python package named modelspec. ModelSpec is an analysis of alternatives for AI models: it decides which models fit a job from your requirements, sourced benchmarks and real cost, and shows its work.
Claims
- The decision comes from your requirements, every admitted benchmark and real cost. Nobody pays to rank higher.
- When the evidence can't separate models, ModelSpec says so: it returns the tie as kind: "tied" with tie_breakers. It does not return the cheapest model as the answer.
- The same spec on the same snapshot returns the same answer, with a decision ID and a signed snapshot.
- ModelSpec decides per job or role. It does not route or proxy requests: a router can choose among the models it recommends.
- People decide free on the board. Machines use the hosted API and the MCP server with a key.
How to call it
What it costs
- A decision costs 1 credit.
- A fully explained decision costs 2 credits.
- A ranking costs 1 credit.
- A licence and data-residency check costs 5 credits.
- Solo costs $29 a month for 2,500 credits.
- Team costs $199 a month for 25,000 credits.
- Scale costs $799 a month for 150,000 credits.
- 250-credit pack costs $5 for 250 credits.
- 1,300-credit pack costs $25 for 1,300 credits.
- 2,750-credit pack costs $50 for 2,750 credits.
- 6,000-credit pack costs $100 for 6,000 credits.