The entire modhub dataset — 51 models, 7 benchmarks, all scores and pricing — as clean JSON. Free, no API key, CORS enabled, responses cached for an hour. Build dashboards, alerts, routers or research on top of it. If you ship something public, a link back is appreciated.
GET/api/v1/models
All models with full specs and pricing. Filter with ?provider=, ?open_weights=true, ?reasoning=true.
curl https://modhub.ai/api/v1/models?open_weights=true
{
"updatedAt": "2026-06-10",
"count": 15,
"models": [
{
"slug": "deepseek-v4",
"name": "DeepSeek V4",
"provider": "deepseek",
"inputPrice": 0.3,
"outputPrice": 0.5,
"contextWindow": 1000000,
"openWeights": true,
"license": "MIT",
...
}
]
}GET/api/v1/models/{slug}
One model, including its benchmark scores and per-benchmark rank.
curl https://modhub.ai/api/v1/models/claude-opus-4-8
{
"model": { "slug": "claude-opus-4-8", ... },
"scores": [
{
"benchmark": "swe-bench-verified",
"score": 88.6,
"isEstimate": false,
"rank": 2
}
]
}GET/api/v1/benchmarks
All tracked benchmarks with full leaderboards.
curl https://modhub.ai/api/v1/benchmarks
{
"benchmarks": [
{
"slug": "swe-bench-verified",
"name": "SWE-bench Verified",
"leaderboard": [
{ "model": "claude-fable-5", "score": 95.0, "isEstimate": true },
...
]
}
]
}GET/api/v1/compare?models=a,b
Head-to-head data for any two models: shared benchmarks, spec deltas, price ratio.
curl "https://modhub.ai/api/v1/compare?models=gpt-5-5,claude-opus-4-8"
{
"a": { "slug": "gpt-5-5", ... },
"b": { "slug": "claude-opus-4-8", ... },
"sharedBenchmarks": [ ... ],
"wins": { "gpt-5-5": 1, "claude-opus-4-8": 3 }
}No authentication is required. Please cache responses on your side (data changes at most daily) and keep request rates reasonable. Attribution with a link to modhub.ai is appreciated but not required. Shapes are stable; new fields may be added without notice.