Builder
One plan. Model usage billed at 1.25x provider cost, no markup surprises.
Looking for per-token API prices instead? See the LLM API pricing comparison →
Builder
Most PopularFor AI builders comparing model quality, latency, and cost. Cancel anytime.
- ✓ $10 model usage included every billing period
- ✓ Flexible top-ups: add model-usage funds on demand ($20, $50, $100). Top-ups do not expire.
- ✓ Transparent billing at 1.25x provider cost
- ✓ All AI Ask API access for programmatic evaluation (your subscription does not include provider API credits)
- ✓ See actual and estimated model costs
- ✓ Side-by-side multi-model comparison
- ✓ Automated AI Judge scoring and evaluation
- ✓ Higher rate limits (60 requests/min)
What does the included usage cover?
The subscription is $20 per month and includes $10 of model usage. Provider token cost is multiplied by 1.25x for usage billing; the subscription price and token price are separate things.
- 500 short comparisons across 3 models is about $0.08 in model usage using text-embedding-3-small as the registry example. Actual cost varies with model, input, output, and search usage.
- 100 deeper evaluations across 3 models is about $0.68 in model usage using gpt-oss-safeguard-20b as the registry example. Actual cost varies with model, input, output, and search usage.
Batch 42 evidence surface · verified 2026-08-27 · exact route allowlist: /pricing
Subscription and tier architecture decision evidence
Batch 42 · M1: Plan-component and source-of-truth ledger
Formula: Sellable plan = application plan ∩ billing configuration; an unjoined field is Unavailable, never inferred from UI copy.
Provenance: Server-readable PLANS and billing-facing display fields joined for the Builder fixture; secret checkout identifiers are intentionally omitted. Verified 2026-08-27.
First-party source: All AI Ask pricing
| Field ID / fixture | Frozen inputs | Observation | Decision boundary | State |
|---|---|---|---|---|
batch42-pricing-m1-r1Builder monthly allowance | plan=builder; fee=$20; interval=monthly; included=10 | Subscription fee and included model-usage balance are separate ledger fields. | Do not call included usage provider API credit or transfer it to another account. | JOINED — display configuration |
batch42-pricing-m1-r2Provider-cost settlement | registry usage × markup; multiplier=1.25x; API=true | Usage charge = provider registry cost × 1.25; external provider credit = excluded. | No charge is calculated without the exact registry rate and usage export. | CALCULATED — formula closed |
batch42-pricing-m1-r3Top-up and checkout identity | denominations=20, 50, 100 USD; reset=monthly; top-up expiry=none | Top-ups are non-expiring model-usage balance; product/price identifiers are Unavailable — secret billing identifiers are not exposed | Never render an unavailable price, tax state, or checkout ID as a commercial value. | Unavailable — checkout join intentionally withheld |
Batch 42 · M2: Frozen workload-to-balance settlement table
Formula: Usage draw = (input units × input rate + output units × output rate + search units × search rate) × markup; remaining = allowance − draw, then top-up.
Provenance: Three frozen evaluation shapes are estimates, not quotes: short comparison, long-context comparison, and tool/search-assisted evaluation. Registry rate date and exact model snapshot are required. Verified 2026-08-27.
First-party source: All AI Ask cost methodology
| Field ID / fixture | Frozen inputs | Observation | Decision boundary | State |
|---|---|---|---|---|
batch42-pricing-m2-r1Short comparison | 3 models; 2,000 input + 500 output each; registry rates dated | Draw = 3 × (2,000×input + 500×output) / 1,000 × 1.25; exact amount Unavailable — matched model-rate export is absent | Estimate cannot rank models or become a quote. | Unavailable — rate join absent |
batch42-pricing-m2-r2Long-context comparison | 3 models; 12,000 input + 3,000 output each; allowance opening=$10 | Allowance draw and remaining balance are Unavailable — snapshot-specific registry usage is absent | Subscription balance is not the same as an external provider balance. | Unavailable — usage settlement absent |
batch42-pricing-m2-r3Tool/search-assisted evaluation | tool output + search units; top-up denomination=$20; provider account excluded | Top-up draw order and uncovered amount are Unavailable — search-unit tariff and invoice reconciliation are absent | No unsupported search cost or external-provider credit may render as zero. | Unavailable — search tariff unavailable |
Batch 42 · M3: Billing lifecycle and entitlement state machine
Formula: Effective entitlement = authenticated state ∧ subscription status ∧ billing event order; unknown tax, grace, refund, or proration remains Unavailable.
Provenance: Frozen anonymous, checkout, trial, active, failed, cancellation, renewal, and refund-requested fixtures. Billing portal action is distinguished from model-usage balance. Verified 2026-08-27.
First-party source: All AI Ask billing terms
| Field ID / fixture | Frozen inputs | Observation | Decision boundary | State |
|---|---|---|---|---|
batch42-pricing-m3-r1Anonymous → checkout-pending | no account; subscribe intent; pending event; no entitlement | Access state is registration-required; checkout outcome Unavailable — provider event is not present | A CTA click never means active subscription. | Unavailable — checkout event unavailable |
batch42-pricing-m3-r2Active → payment-failed → recovery | status sequence; API entitlement; rate limit; monthly reset; top-up balance | Recovery path and grace-period duration are Unavailable — billing policy join is absent | Do not infer grace, refund, tax, or proration behavior. | Unavailable — grace policy unavailable |
batch42-pricing-m3-r3Cancellation → renewed / refund-requested | effective date; portal action; allowance reset; preserved top-up; event order | Top-up preservation is recorded separately; refund eligibility is Unavailable — case-specific billing review is absent | Cancellation and refund are not synonymous; transactional state stays out of SEO claims. | Unavailable — refund decision unavailable |
Decision boundary: unresolved identity, control, usage, quality, parity, tariff, entitlement, or lifecycle fields remain Unavailable; they never become zero, supported, passing, active, or equivalent.
Test your subscription evaluation path →Pricing questions
Can I cancel?
Yes. Cancel anytime from the billing portal. Access and remaining plan usage follow the subscription status shown there; unused monthly allowance does not roll over.
Are top-ups included?
No. Top-ups are optional one-time purchases in the available amounts above. They are separate from the monthly subscription and do not expire.
Does this include OpenAI, Anthropic, or other provider API credits?
No. All AI Ask API access lets you use this service programmatically. Provider API accounts and their per-token bills are separate. See the LLM API pricing comparison for provider token rates.