The ultimate usage dashboard plugin for your AI coding assistants.
OpenCode Quotas is an OpenCode.ai plugin that aggregates usage data from Antigravity, Codex, and GitHub Copilot into a single, beautiful dashboard injected directly into your OpenCode chat footer. Never hit a rate limit unexpectedly again.
Note: This is a community-developed plugin and is not officially affiliated with OpenCode.ai.
- Unified Dashboard: See all your AI quotas (Antigravity, Codex, Copilot) in one place.
- Smart Predictions: Uses linear regression on usage history to predict when you'll exhaust your quota.
- Visual Progress Bars: ANSI-colored bars that change from green to yellow to red as you approach limits.
- Context Aware: Optionally filter quotas to show only those relevant to the active model.
- Resilient: Provider failures are isolated and won't break your chat experience.
Clone the repository into your OpenCode plugins directory and build it:
# Clone into your plugins directory
git clone https://github.com/your-org/opencode-quotas ~/.opencode/plugins/opencode-quotas
# Build the plugin
cd ~/.opencode/plugins/opencode-quotas
bun install
npm run buildAdd opencode-quotas to your OpenCode configuration file (typically ~/.opencode/config.json) under the plugins array:
{
"plugin": [
"opencode-antigravity-auth",
"opencode-quotas"
]
}Once installed and registered, a live quota summary is automatically appended to the final assistant response footer in OpenCode. The plugin waits for the session to go idle before injecting to avoid duplicate footers during multi-step runs.
You can also use OpenCode Quotas as a standalone CLI tool to check quotas directly from your terminal:
# Show all quotas
opencode-quotas
# Filter by model (simulates what the plugin does automatically)
opencode-quotas --provider google --model antigravity-gemini-3-flashTip: During development, use
bun run opencode-quotasto run without building.
| Provider | Raw Quota IDs | Aggregated IDs | Description |
|---|---|---|---|
| Antigravity | ag-raw-* |
ag-flash, ag-pro, ag-premium |
Raw model quotas aggregated by pattern into Flash, Pro, and Premium tiers. |
| Codex | codex-primary, codex-secondary |
codex-smart |
Rate limits aggregated using most-critical strategy. |
| GitHub Copilot | — | — | enableExperimentalGithub: true in config, disabled by default) |
Note: Quota IDs are used in configuration options like
disabled.
| Column | Description |
|---|---|
status |
Health indicator: OK (normal), WRN (warning), ERR (at/over limit) |
name |
Display name of the quota (e.g., "Antigravity Flash") |
bar |
Visual progress bar showing utilization |
percent |
Numeric percentage of quota used |
reset |
Time until the quota resets (e.g., "23m", "1h 30m") |
ettl |
Estimated Time To Limit - predicted time until quota exhaustion based on usage trends |
value |
Raw used/limit values (e.g., "150/500 credits") |
window |
Rate limit window duration (e.g., "5h window") |
info |
Additional info or alerts (e.g., "unlimited", "!!") |
All configuration is done via .opencode/quotas.json in your project root. Settings apply to both the CLI and plugin.
By default, filterByCurrentModel is disabled (false). Enable it to show only quotas relevant to the active model.
{
"filterByCurrentModel": false
}How it works: The plugin extracts tokens from the model ID (e.g., antigravity-gemini-3-flash becomes ["antigravity", "gemini", "3", "flash"]) and scores each quota by how many tokens match its ID or name. The highest-scoring quotas are displayed.
Before filtering (all quotas):
ST QUOTA NAME USED UTILIZATION RESET ETTL
--- ------------------- ---- -------------------- ------ -----
OK Antigravity Flash 20% ████░░░░░░░░░░░░░░░░ 28m 13m
OK Antigravity Premium 40% ████████░░░░░░░░░░░░ 56m 1h 0m
ERR Codex Usage 100% ████████████████████ 13h 9m -
After filtering (model: google/antigravity-gemini-3-flash):
ST QUOTA NAME USED UTILIZATION RESET ETTL
--- ----------------- ---- -------------------- ----- ----
OK Antigravity Flash 20% ████░░░░░░░░░░░░░░░░ 28m 13m
Permanently hide quotas you don't care about:
{
"disabled": ["ag-pro", "codex-smart"]
}Combine multiple quotas into a single display row using explicit sources or pattern matching.
Pattern-based aggregation (matches raw quota IDs/names):
{
"aggregatedGroups": [
{
"id": "ag-flash",
"name": "Antigravity Flash",
"patterns": ["flash"],
"providerId": "antigravity",
"strategy": "most_critical"
}
]
}Explicit source aggregation (specify exact quota IDs):
{
"aggregatedGroups": [
{
"id": "codex-unified",
"name": "Codex Usage",
"sources": ["codex-primary", "codex-secondary"],
"strategy": "most_critical"
}
]
}Aggregation options:
| Option | Type | Description |
|---|---|---|
id |
string | Unique identifier for the aggregated quota |
name |
string | Display name |
sources |
string[] | Explicit quota IDs to include |
patterns |
string[] | Patterns to match against raw quota IDs/names |
providerId |
string | Limit pattern matching to a specific provider |
strategy |
string | Aggregation strategy (see below) |
Aggregation strategies:
| Strategy | Description |
|---|---|
most_critical |
Shows the quota predicted to hit its limit soonest (requires usage history) |
max |
Shows the quota with highest current usage percentage |
min |
Shows the quota with lowest current usage percentage |
mean |
Displays average usage across all sources |
median |
Displays median usage across all sources |
All options below work in both CLI and plugin output. ANSI colors only render in terminal environments.
{
"table": {
"columns": ["name", "bar", "percent"]
}
}QUOTA NAME UTILIZATION USED
------------------- -------------------- ----
Antigravity Flash ████░░░░░░░░░░░░░░░░ 20%
Antigravity Premium ████████░░░░░░░░░░░░ 40%
Codex Usage ████████████████████ 100%
{
"progressBar": {
"filledChar": "=",
"emptyChar": "-",
"width": 10
}
}ST QUOTA NAME USED UTILIZATION RESET
--- ------------------- ---- ----------- ------
OK Antigravity Flash 20% ==-------- 23m
OK Antigravity Premium 40% ====------ 51m
ERR Codex Usage 100% ========== 13h 4m
Switch from "used" (default) to "available" mode. This inverts the display—useful if you prefer seeing how much is left:
{
"progressBar": {
"show": "available"
}
}ST QUOTA NAME USED UTILIZATION RESET
--- ------------------- ---- -------------------- ------
WRN Antigravity Flash 80% ████████████████░░░░ 23m
OK Codex Usage 0% ░░░░░░░░░░░░░░░░░░░░ 13h 4m
Note: In "available" mode, the percentage shows remaining capacity, and status thresholds are inverted.
{
"table": {
"header": false
}
}OK Antigravity Flash 20% ████░░░░░░░░░░░░░░░░ 23m 5h 52m
OK Antigravity Premium 40% ████████░░░░░░░░░░░░ 51m 1h 2m
ERR Codex Usage 100% ████████████████████ 13h 4m -
For plain text output (logging, CI, or non-TTY environments):
{
"progressBar": {
"color": false
}
}Define when colors change based on usage percentage:
{
"progressBar": {
"color": true,
"gradients": [
{ "threshold": 0.5, "color": "green" },
{ "threshold": 0.8, "color": "yellow" },
{ "threshold": 1.0, "color": "red" }
]
}
}Available colors: red, green, yellow, blue, magenta, cyan, white, gray
See schemas/quotas.schema.json for the complete JSON Schema.
| Option | Type | Default | Description |
|---|---|---|---|
footer |
boolean | true |
Show quotas in chat footer |
showFooterTitle |
boolean | true |
Show plugin title/header in footer |
filterByCurrentModel |
boolean | false |
Filter quotas by active model |
disabled |
string[] | [] |
Quota IDs to hide |
showUnaggregated |
boolean | true |
Show quotas that don't match any aggregation group |
progressBar.color |
boolean | false |
Enable ANSI colors (set true for colored terminal output) |
progressBar.width |
number | 20 |
Progress bar character width |
progressBar.show |
string | "used" |
"used" or "available" |
table.columns |
string[] | (auto) | Columns to display |
table.header |
boolean | true |
Show column headers |
aggregatedGroups |
array | (see defaults) | Quota aggregation with patterns or sources |
historyMaxAgeHours |
number | 24 |
Max history age in hours |
pollingInterval |
number | 60000 |
Refresh interval in ms |
predictionWindowMinutes |
number | 60 |
Time window for regression analysis (minutes) |
predictionShortWindowMinutes |
number | 5 |
Short time window for spike detection (minutes) |
debug |
boolean | false |
Enable debug logging |
Your credentials remain safe. This plugin uses standard local OAuth flows and stores tokens securely on your machine (~/.config/opencode/). No data is sent to third-party servers other than the quota providers themselves.
Built with Bun for speed.
bun install # Install dependencies
bun test # Run tests
OPENCODE_QUOTAS_E2E=1 bun test tests/integration/injection-e2e.test.ts # Run E2E injection test
npm run build # Build for production
bun run opencode-quotas # Run CLI in dev modeMIT