Back to the service list
Hosted service / Terminal / IDE

Command Code

A coding agent built around open models. The per-model spend allowance and API access differ by plan.

Platforms: Windows / macOS / Linux / CLI

Plans and what they cover

Not disclosed means the official page states no figure; not researched means we have not checked the item yet. Neither is the same as "no limit".

Go

Subscription + usage

$1.00 / month

Tax treatment not researched. Payment fee separate

$1 per month plus a payment fee. Go does not include API or cloud access. The per-model spend allowances are drawn from one shared pool, so the figures for different models cannot be added together. Only some models are listed.

Models and their allowances

Listed models are those we could confirm. When a model has no individual figure, check the shared plan allowance or separate contract terms as well.

  • GLM-5.3-FlashAvailable

    Included in Go. The $10 per month equivalent is what this model alone would use; it is not a separate pot to add on top of other models.

    • Monthly spend equivalent per model, used alone: 10 USD equivalent / month

      The official estimate is 1,770 requests per 5 hours, 3,540 per week and 5,900 per month. It assumes about 800 input, 50,000 cache-read and 125-200 output tokens per request, and is not a guaranteed count.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Without extra credits, paid models stop

      Official source for the per-model allowance
  • GPT 6 LunaAvailable

    Included in Go. The $6 per month equivalent is what this model alone would use; it is not a separate pot to add on top of other models.

    • Monthly spend equivalent per model, used alone: 6 USD equivalent / month

      The official estimate is 2,730 requests per 5 hours, 5,450 per week and 9,090 per month. It assumes about 800 input, 50,000 cache-read and 125-200 output tokens per request, and is not a guaranteed count.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Without extra credits, paid models stop

      Official source for the per-model allowance
  • Qwen 3.8 27BAvailable

    Included in Go. The $10 per month equivalent is what this model alone would use.

    • Monthly spend equivalent per model, used alone: 10 USD equivalent / month

      The official estimate is 1,030 requests per 5 hours, 2,050 per week and 3,420 per month. It assumes about 800 input, 50,000 cache-read and 125-200 output tokens per request, and is not a guaranteed count.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Without extra credits, paid models stop

      Official source for the per-model allowance

Usage limit details

Shared allowance10 USD equivalent / month
Scope
Whole plan
Window
month
Resets
At the start of the billing cycle. Also has 5-hour ($3 equivalent) and 7-day ($6 equivalent) windows
On overage
Buy extra credits or auto top-up (optional). Without them, paid models stop until the allowance resets
Rollover
Bought credits do not roll over or expire. The subscription portion has not been researched

Using several models draws on a shared pool in proportion. Each model has a different spend equivalent, so the listed figures for different models cannot be added together.

Official source for this limit

Officially confirmed · Checked

Official plan page for the latest terms
Official sources read (2)

GOAT

Subscription + usage

$10.00 / month

Tax treatment not researched. Payment fee separate

$10 per month plus a payment fee. The spend equivalent is not a cash balance and differs per model; it does not mean the full amount works with every model. Only some supported models are listed.

Models and their allowances

Listed models are those we could confirm. When a model has no individual figure, check the shared plan allowance or separate contract terms as well.

  • GLM-5.3-FlashAvailable

    Listed officially. The spend equivalent and request count depend on the model and token volume.

    • Per-model spend equivalent cap: 100 USD equivalent / month

      The official estimate is 11,800 requests per 5 hours, 29,500 per week and 59,000 per month. It is a guide based on a typical case of about 800 input, 50,000 cache-read and 125-200 output tokens per request, not a guarantee.

      Resets: Applies 5-hour, 7-day and billing-cycle windows. Check officially for exact per-model figures / On overage: With extra credits it is usage-based; without them, paid models stop until the reset

      Official source for the per-model allowance
  • GPT 6 LunaAvailable

    Listed officially. The spend equivalent and request count depend on the model and token volume.

    • Per-model spend equivalent cap: 20 USD equivalent / month

      The official estimate is 6,060 requests per 5 hours, 15,200 per week and 30,300 per month. It is a guide based on a typical case of about 800 input, 50,000 cache-read and 125-200 output tokens per request, not a guarantee.

      Resets: Applies 5-hour, 7-day and billing-cycle windows. Check officially for exact per-model figures / On overage: With extra credits it is usage-based; without them, paid models stop until the reset

      Official source for the per-model allowance

Usage limit details

Plan-wide limits have not been researched. Per-model allowances are shown above.

Officially confirmed · Checked

Official plan page for the latest terms
Official sources read (2)

Pro

Subscription + usage

$20.00 / month

Tax treatment not researched. Payment fee separate

$20 per month plus a payment fee. Claude Opus and Fable are excluded. API and cloud access are included. Per-model, premium and total allowances coexist, and the official page explains when each is drawn down. Only some models are listed.

Models and their allowances

Listed models are those we could confirm. When a model has no individual figure, check the shared plan allowance or separate contract terms as well.

  • GLM-5.3-FlashAvailable

    Included in Pro. The $110 per month equivalent is what this model alone would use.

    • Monthly spend equivalent per model, used alone: 110 USD equivalent / month

      The official estimate is 13,000 requests per 5 hours, 32,400 per week and 64,900 per month. It assumes about 800 input, 50,000 cache-read and 125-200 output tokens per request, and is not a guaranteed value.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Stops without extra credits

      Official source for the per-model allowance
  • GPT 6 LunaAvailable

    Included in Pro. The $30 per month equivalent is what this model alone would use.

    • Monthly spend equivalent per model, used alone: 30 USD equivalent / month

      The official estimate is 9,090 requests per 5 hours, 22,700 per week and 45,500 per month. It assumes about 800 input, 50,000 cache-read and 125-200 output tokens per request, and is not a guaranteed value.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Stops without extra credits

      Official source for the per-model allowance
  • Claude Sonnet 5.5Available

    Included in Pro. Premium models are capped at $20 per month each, assessed separately from the $110 shared allowance.

    • Monthly spend equivalent per premium model: 20 USD equivalent / month

      The official estimate is 476 requests per 5 hours, 1,190 per week and 2,380 per month. It assumes about 800 input, 50,000 cache-read and 125-200 output tokens per request, and is not a guaranteed value.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Stops without extra credits

      Official source for the per-model allowance
  • Claude Haiku 5.5Available

    Included in Pro, $30 per month equivalent. Check the official list for per-version terms.

    • Monthly spend equivalent per premium model: 30 USD equivalent / month

      The official estimate is 8,960 requests per 5 hours, 22,400 per week and 44,800 per month. It varies with the model and token counts and is not a guaranteed value.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Stops without extra credits

      Official source for the per-model allowance
  • Claude Opus (family)Not available

    Officially, Claude Opus requires the Max plan or usage-based Provider API billing.

    This model’s own allowance: not researched, or shared across the plan. See the usage limits below.

  • Qwen 3.8 MaxAvailable

    Included in Pro. The $30 per month equivalent is what this model alone would use.

    • Monthly spend equivalent per model, used alone: 30 USD equivalent / month

      The official estimate is 392 requests per 5 hours, 980 per week and 1,960 per month. It assumes about 800 input, 50,000 cache-read and 125-200 output tokens per request, and is not a guaranteed value.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Stops without extra credits

      Official source for the per-model allowance
  • MiniMax M3Available

    Included in Pro, $57 per month equivalent. Calculated at the official 50% discount.

    • Monthly spend equivalent per model, used alone: 57 USD equivalent / month

      The official estimate is 3,360 requests per 5 hours, 8,410 per week and 16,800 per month. The listed price is the limited-time 50% discount. Check the official page for the end date and the normal price.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Stops without extra credits

      Official source for the per-model allowance

Usage limit details

Total allowance110 USD equivalent / month
Scope
Whole plan
Window
month
Resets
At the start of the billing cycle. Also has 5-hour ($22 equivalent) and 7-day ($55 equivalent) windows
On overage
With extra credits it is usage-based; without them, paid models stop while free models keep working
Rollover
Bought credits do not roll over or expire. The subscription portion has not been researched

The official figures are $22 per 5 hours, $55 per 7 days and $110 per month. Per-model allowances apply separately from this total, so the total cannot be freely divided between models.

Official source for this limit

Officially confirmed · Checked

Official plan page for the latest terms
Official sources read (2)

Max 10×

Subscription + usage

$100.00 / month

Tax treatment not researched. Payment fee separate

$100 per month plus a payment fee. Only some models are listed. Standard and premium have separate allowances, and the official ~230K requests estimate assumes a 43% standard / 57% premium mix.

Models and their allowances

Listed models are those we could confirm. When a model has no individual figure, check the shared plan allowance or separate contract terms as well.

  • Claude Opus (family)Available

    Max covers all 100 models including Claude Opus. Check the official list for per-version terms.

    • Monthly spend equivalent per premium model: 100 USD equivalent / month

      The official estimate for Opus 5.5 is 1,790 requests per 5 hours, 3,570 per week and 5,950 per month; for Opus 5 and 4.8 it is 896 per 5 hours, 1,790 per week and 2,990 per month. Prices and counts differ by version.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Check the official page for when standard allowance and extra credits apply

      Official source for the per-model allowance
  • GLM-5.3-FlashAvailable

    Counts against the $150 standard-model allowance.

    • Monthly spend equivalent per standard model: 150 USD equivalent / month

      The official estimate is 26,500 requests per 5 hours, 53,100 per week and 88,500 per month. It is a guide based on typical token counts, not a guaranteed value.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Stops without extra credits

      Official source for the per-model allowance
  • Claude Sonnet 5.5Available

    Billed in the standard model class and counts against the $150 standard allowance.

    This model’s own allowance: not researched, or shared across the plan. See the usage limits below.

  • Claude Haiku 5.5Available

    Billed in the standard model class and counts against the $150 standard allowance.

    This model’s own allowance: not researched, or shared across the plan. See the usage limits below.

  • GPT 6 LunaAvailable

    Advertised as premium, but the actual billing class is standard. Counts against the $150 standard allowance.

    • Monthly spend equivalent per standard model: 150 USD equivalent / month

      The official estimate is 68,200 requests per 5 hours, 136,000 per week and 227,000 per month. It is presented as premium but billed in the standard class.

      Resets: Follows the shared 5-hour, 7-day and monthly windows / On overage: Stops without extra credits

      Official source for the per-model allowance
  • Qwen 3.8 MaxAvailable

    Counts against the $150 standard-model allowance.

    This model’s own allowance: not researched, or shared across the plan. See the usage limits below.

Usage limit details

Standard model allowance150 USD equivalent / month
Scope
Whole plan
Window
month
Resets
At the start of the billing cycle. Also has 5-hour ($45 equivalent) and 7-day ($90 equivalent) windows
On overage
Extra credits work with every model, with no cap
Rollover
Bought credits do not roll over or expire. The subscription portion has not been researched

The monthly allowance for standard models: $45 per 5 hours and $90 per 7 days. Premium-class models get a separate $100 allowance, so the $250 total does not simply sit in one wallet.

Official source for this limit
Premium model allowance100 USD equivalent / month
Scope
Whole plan
Window
month
Resets
At the start of the billing cycle
On overage
Extra credits work with every model
Rollover
Bought credits do not roll over or expire. The subscription portion has not been researched

For the premium billing class, such as Claude Opus and Fable. The billing class, not the brand name, decides whether usage comes from the standard or premium allowance.

Official source for this limit

Officially confirmed · Checked

Official plan page for the latest terms
Official sources read (1)

Max 20×

Subscription + usage

$200.00 / month

Tax treatment not researched. Payment fee separate

$200 per month plus a payment fee. The official ~370K requests estimate assumes a 43% standard / 57% premium mix. Only some models are listed.

Models and their allowances

Listed models are those we could confirm. When a model has no individual figure, check the shared plan allowance or separate contract terms as well.

  • Claude Opus (family)Available

    Max covers all 100 models including Claude Opus.

    This model’s own allowance: not researched, or shared across the plan. See the usage limits below.

  • GLM-5.3-FlashAvailable

    Counts against the $300 standard-model allowance.

    This model’s own allowance: not researched, or shared across the plan. See the usage limits below.

  • GPT 6 LunaAvailable

    Advertised as premium, but the billing class is standard.

    This model’s own allowance: not researched, or shared across the plan. See the usage limits below.

Usage limit details

Standard model allowance300 USD equivalent / month
Scope
Whole plan
Window
month
Resets
At the start of the billing cycle. Also has 5-hour ($90 equivalent) and 7-day ($180 equivalent) windows
On overage
Extra credits work with every model, with no cap
Rollover
Bought credits do not roll over or expire. The subscription portion has not been researched

Twice Max 10x: $90 per 5 hours and $180 per 7 days. The $200 premium allowance applies separately.

Official source for this limit
Premium model allowance200 USD equivalent / month
Scope
Whole plan
Window
month
Resets
At the start of the billing cycle
On overage
Extra credits work with every model
Rollover
Bought credits do not roll over or expire. The subscription portion has not been researched

For the premium billing class, such as Claude Opus and Fable. Assessed separately from the standard allowance.

Official source for this limit

Officially confirmed · Checked

Official plan page for the latest terms
Official sources read (1)

Sources for this service