PUBLIC / LIVE
Scheduling overview
The honest observable surface of the scheduling layer: platform probe, three-state model availability, dependency checks. Where there is no data we say unavailable.
Platform probe
ok
2026-11-07T02:05:14.932657Z
Verified models
0
24 models in total
Status unknown
24
24 models in total
Needs review
0
24 models in total
Dependency checks
| Check | State |
|---|---|
| db | ok |
| audit_key | ok |
Metrics not wired yet
- Requests per second, active nodes, GPU hours and live prices need the scheduling scheduling telemetry, which is not open yet.
- No invented curve or “sample data” is passed off as monitoring here; once a monitor is wired the values are replaced by server responses.
- This page is never cached (no-store): what you see is what the probe just returned.
Data source and version
Page generated at 2026-11-07T02:05:14Z
Source /healthz · Contract version 1.2.5
Source /v1/models · Contract version 1.2.5
Model availability
Public model list (server-provided; no model count is inferred when the service is unavailable)
| Model | Owner | Context window | Tokenizer | Reference price | Status |
|---|---|---|---|---|---|
| claude-fable-5-1 | djb | not synced | — | Status unknown | |
| claude-opus-5 | djb | not synced | — | Status unknown | |
| claude-opus-5-5 | djb | not synced | — | Status unknown | |
| claude-sonnet-5 | djb | not synced | — | Status unknown | |
| deepseek-v4-flash | djb | not synced | — | Status unknown | |
| deepseek-v4-pro | djb | not synced | — | Status unknown | |
| gemini-3.1-flash-image-preview | djb | not synced | — | Status unknown | |
| gemini-3.6-flash | djb | not synced | — | Status unknown | |
| glm-5.3 | djb | not synced | — | Status unknown | |
| glm-5.3-flash | djb | not synced | — | Status unknown | |
| gpt-5.6-luna | djb | not synced | — | Status unknown | |
| gpt-5.6-sol | djb | not synced | — | Status unknown | |
| gpt-5.6-terra | djb | not synced | — | Status unknown | |
| gpt-6-astra | djb | not synced | — | Status unknown | |
| gpt-6-luna | djb | not synced | — | Status unknown | |
| gpt-6-sol | djb | not synced | — | Status unknown | |
| gpt-6.1-sol | djb | not synced | — | Status unknown | |
| gpt-image-2 | djb | not synced | — | Status unknown | |
| gpt-image-2.5-flare | djb | not synced | — | Status unknown | |
| gpt-image-2.5-sunburst | djb | not synced | — | Status unknown | |
| kimi-k3 | djb | not synced | — | Status unknown | |
| kimi-k3.1-pro | djb | not synced | — | Status unknown | |
| qwen3.7-max | djb | not synced | — | Status unknown | |
| qwen3.8-max | djb | not synced | — | Status unknown |
Status comes from the service probe. “Unknown” means the service did not return an availability field for that model and we do not infer one; when there is no data we say “unavailable”, never 0.