Output tokens by provider and model
Registered token prices
Cache, input, and output rates used to price ledger turns. Empty cache falls back to input. Save writes ~/.rtxclaw/pricing.json.
| Provider | Model | Input | Cache | Cache write | Output | |
|---|---|---|---|---|---|---|
| No pricing table loaded | ||||||
Generation logs
| When | Provider | Agent | Model | In | Cached | Out | TTFT | Gen | Wait | Cost |
|---|---|---|---|---|---|---|---|---|---|---|
| No generations recorded | ||||||||||
Local cluster
—
vLLM mux · GPUs, models, fallbacks, and capacity per server
—Servers
—Replicas up
—Queries in flight
—Concurrent cap
Routing models
Auto-loads the current operator vllm-router.yaml.
Add, edit, or delete a model to update that list; live cards refresh after save.
Subscriptions
Live remaining quota and prepaid API credits for providers discovered on this machine. Each card appears immediately and fills in as that vendor answers.
No subscription snapshot yet