Waiting for live-lane analysis.
What works, what runs out, what to unlock next.
Real state first: live lanes, budget posture, data freshness, and the next useful account to unlock.
Waiting for budget and freshness state.
Can this actually run?
Building computed portfolio state from current artifacts.
Waiting for provider ranking.
Waiting for artifact timestamps.
Waiting for budget and lane checks.
Route board
Recent burn and change
No change observed yet.
Waiting for a comparison between current and prior runs.
What is usable right now
Right now this is the only confirmed live inference lane from this machine.
Free-model pressure is mostly request-shaped: per-minute, per-day, and sometimes account-state mediated.
Small failures amplify into stalls when retries, request caps, and sparse coverage compound.
Waiting for current OpenRouter status.
This comes from the live `/api/v1/key` response.
Request-based wall, not token-based wall.
How the catalog breaks down
Cut the list down
Observed tokens and latency
Waiting for enough history to classify pressure.
Based on successful live probes in stored history.
Need more live providers before real routing decisions stabilize.
Reset and limit notes
What to get next
One by one
Reliable table
| Provider | Status | Access | Horizon | Worth | Latency | Blocker / note |
|---|