Models
Browse all available models, their pricing, and real-time status.
204
Total Models
13
Free
Throttled rate limits
191
All Plans
Available on every paid tier and Pay2Go
RPM = Requests per Minute · RPD = Requests per Day
Prices are shown separately for input and output per 1 million tokens.
1K = 1,000 (thousand) · 1M = 1,000,000 (million)
Operational
Last probe succeeded with normal latency, or ≥ 80% of recent live traffic succeeded.
Degraded
Latency above the threshold (≥ 10s by default), or 50–80% of recent live traffic succeeded.
Partial Outage
Just transitioned between up and down, or 20–50% of recent live traffic succeeded.
Major Outage
Two consecutive failed probes (~10 min), or under 20% of recent live traffic succeeded.
Probes run every 5 minutes. Live request outcomes (5xx, 429, connection errors) override the probe within a 20-request rolling window for higher accuracy.
Free models & rate limits
Free models can return a 429 “rate limit exceeded” error even when the badge is Operational. That's the per-plan throughput cap (requests per minute / day) — not a model outage. Subscribe or top up your pay-as-you-go balance to lift the limit.
Audio, video & rate-sensitive models
Some routes (text-to-speech, music, voice cloning, video, dubbing, and a few rate-sensitive free chat models) can't be safely probed without burning quota. They show as Operational by default; their true availability appears in the 7-day uptime bars once real traffic flows through them.