What models the free tier is actually running.
Live aggregate of every model FreeRide has routed through OpenRouter across the last 30 days. Updated every 6 hours from OR's public app activity, mirrored here at api.free-ride.xyz/v1/stats.
Tokens per day — last 7 days
trend · oldest → newestTokens by model vendor — last 30 days
share across the top-10 setBars are scaled to the leading vendor. All routing in this set goes through OpenRouter — that's the only provider where we currently capture per-model attribution. Per-provider beacon telemetry for Groq / NVIDIA / Cerebras / HF is on the roadmap.
Top 10 models — last 30 days
ranked by tokens served| # | model | via | tokens | share |
|---|---|---|---|---|
| 01 | openrouter/owl-alpha | OpenRouter (native) | 1,112,556,094 | 79.6% |
| 02 | nvidia/nemotron-3-super-120b-a12b-20230311 | NVIDIA · via OR | 174,655,976 | 12.5% |
| 03 | openai/gpt-oss-120b | OpenAI OSS · via OR | 27,252,703 | 1.9% |
| 04 | nex-agi/nex-n2-pro | nex-agi | 25,330,329 | 1.8% |
| 05 | nvidia/nemotron-3-nano-30b-a3b | NVIDIA · via OR | 13,490,607 | 1.0% |
| 06 | poolside/laguna-m.1-20260312 | Poolside · via OR | 10,070,567 | 0.7% |
| 07 | nvidia/nemotron-3-nano-omni-30b-a3b-reasoning-20260428 | NVIDIA · via OR | 9,816,289 | 0.7% |
| 08 | poolside/laguna-xs.2-20260421 | Poolside · via OR | 9,638,448 | 0.7% |
| 09 | Others | others | 8,323,738 | 0.6% |
| 10 | nvidia/nemotron-nano-12b-v2-vl | NVIDIA · via OR | 7,399,486 | 0.5% |
Share is relative to the top-10 set, not all 30-day traffic.
Why this exists
FreeRide's gateway also reads this leaderboard when an agent asks for model: "auto". The resolver scores each candidate model by failover headroom plus a soft popularity bonus from these numbers, so heavily-used, well-supported models surface first across every install. Your traffic, fed back into routing decisions for everyone else.