Replicate
A finite allowance for new accounts; it does not recur.
https://api.replicate.com/v1Models mentioned
The offer covers a dynamic catalog, provider-selected route, or model class without stable model IDs in this snapshot.
Equivalent paid value
How this was valued: The published allowance or a defensible paid comparison is not precise enough to calculate.
The free allowance is a deliberately undisclosed number of runs on the rotating try-for-free collection of media models: Replicate documents the granted-credit rate limit (1 request/second, maximum 6 requests/minute without a payment method) and per-model paid prices, but publishes no credit face value, run count, duration, or reset anywhere in its billing, prepaid-credit, or rate-limit documentation, and its own "free for a limited number of runs" and "after a bit you'll be asked to set up billing" language contradicts any assumed saturation window, so no finite envelope survives the model-spec or modeled tiers.
Limits and terms
- Coverage
- Dynamic subset of models can be run free for a small unquantified allowance.
- Recurrence
- none documented
- No Card Granted Credit Requests Per Second
- 1
- No Card Granted Credit Requests Per Minute
- 6
- Caveat
- Billing setup is required after the model-specific initial allowance is exhausted.
What happens to your prompts?
API data is deleted after one hour by default, but community models, resultant data, and web history introduce additional handling.
- Plan Scope
- Replicate API, web predictions, marketplace models, custom deployments, and training; each model can also impose third-party terms.
- Prompt Retention
- API prediction inputs, outputs, files, and logs are removed after one hour by default. Web-interface prediction data is retained indefinitely until manually deleted.
- Response Retention
- Same API one-hour and web indefinite retention rules. Training artifacts, custom models, and customer-saved files have separate resource lifecycles.
- Ordinary Logging
- Prediction metadata survives removal of input and output data, and Replicate collects service-performance and resultant data for billing, operations, development, diagnostics, and improvement.
- Model Training
- Replicate's Customer Data license permits training customer-requested derivative models but does not expressly authorize training unrelated foundation models on ordinary inference content. Model authors or third-party offerings can add separate terms.
- Product Improvement
- Replicate may compile non-Customer-Data Resultant Data and use it to improve its services and products; the terms limit Customer Data use to service delivery, customer-requested derivative training, and resultant-data creation.
- Human Or Operator Access
- Content is stored during the stated prediction window and may be processed for delivery, support, security, and legal obligations; community model code receives the full input and must be trusted separately.
- Subprocessors And Routing
- Replicate uses listed subprocessors and executes models supplied by Replicate or community authors. Secret fields are redacted after delivery, but a model author can still misuse a secret received by model code.
- Deletion Controls
- API content expires after one hour by default. Web predictions can be manually deleted, which removes their input and output data and files; users must manage training and model resources separately.
- Caveat
- API and web retention differ sharply, and running third-party model code creates a separate trust boundary. Do not assume a model author cannot inspect an input merely because Replicate later deletes it.
Governing documents
- Terms Of Service
- https://replicate.com/terms/
- Privacy Policy
- https://replicate.com/privacy/
Eligibility
- Account Required
- Yes
- Payment Method Required Initially
- No
Primary sources
Before you build with Replicate
Read the classification narrowly
This record has current first-party evidence for some zero-cost hosted inference. The label describes the bounded offer supported by evidence on 2026-09-19; it does not guarantee permanence, production suitability, uptime, latency, model quality, or access from every region.
Resolve the live model route
No stable model ID is recorded. The offer may use a dynamic catalog, a provider-selected route, or a model class, so resolve the current machine-readable ID before writing a fixed production configuration. The recorded base endpoint is https://api.replicate.com/v1, but the provider's current API reference remains authoritative for paths, authentication, and request shape.
Confirm account and billing boundaries
An account is required, so public catalog evidence cannot by itself prove the limits, balance, or billing behavior attached to your workspace. The reviewed evidence explicitly says no payment card is required. Re-check the signup flow because eligibility and billing controls can change after the snapshot. The equivalent paid value shown here prices a documented allowance where possible; it is not cash, guaranteed savings, or protection from overage.
Re-read the prompt policy for your plan
This record classifies the reviewed handling as “Partially private”: API data is deleted after one hour by default, but community models, resultant data, and web history introduce additional handling. Provider policies can distinguish free, paid, enterprise, opted-in, and feature-specific traffic, so verify the governing terms for the exact account and route that will receive your data.
Follow the evidence, then re-check it
The record links 11 first-party sources covering the offer, catalog, limits, pricing, terms, privacy, or adoption evidence available to the audit. Prefer the newest governing document or live catalog when sources disagree, and submit a correction when a provider changes a material term.
Match the quota to the workload shape
Translate the published allowance into the traffic pattern you actually expect instead of comparing headline totals alone. A daily token pool can look generous while a low requests-per-minute or concurrency ceiling blocks interactive bursts; a high request limit can still fail a long-context job when input, output, or per-request tokens are capped. Separate prompt tokens from generated tokens, include retries and tool calls, and test the largest realistic payload. If the provider documents more than one limit window, the tightest window at your peak load is the practical ceiling. Research, experimental, and community access can also carry eligibility or fair-use constraints that cannot be modeled as a simple number.
Plan fallback without changing the rules
A fallback should preserve more than API syntax. Confirm that the substitute route supports the required modality, context length, streaming behavior, structured output, tools, and safety controls, then compare its prompt-retention and training terms. An OpenAI-compatible request shape does not make providers operationally or contractually equivalent. Decide which errors may trigger a retry, cap retry storms, and prevent an exhausted free route from silently switching to a billable model. If deterministic output matters, record the model revision and sampling settings; rotating aliases and free-model pools can change behavior even when the endpoint remains available.
Monitor the offer as a dependency
Capture the model ID, response model field, rate-limit headers, usage fields, latency, HTTP status, and any provider request identifier for each test. Watch for authorization failures, quota exhaustion, catalog removal, policy revisions, and dashboard balance changes as separate failure modes. Re-check the provider’s live catalog and governing pages on a schedule proportionate to the workload’s importance, and keep the dated sources that supported your decision. Free capacity is especially suitable for prototypes, evaluations, fallbacks, and bounded workloads when the application can tolerate change; a production dependency still needs observability, an exit path, and an owner responsible for re-verification.
Test one complete request before scaling
Start with the smallest permitted request using the exact credential, model ID, endpoint, and account type you intend to deploy. Record the HTTP status, response headers, usage fields, latency, and any dashboard balance change. Then exercise the failure path: an invalid model, an exhausted quota, or a rate-limit response should fail clearly without silently switching to a paid route. If the provider supports streaming or tool calls, validate those features separately because a free model can expose a narrower capability set than its paid counterpart. Keep a budget ceiling outside the application whenever billing is possible, and avoid sending sensitive data until the observed route matches the reviewed data agreement.