The shame page

The authoritative current FAQ says the Free Tier is paused; live aliases containing :free do not override that statement.

Current Ai2 deployment documentation describes customer-deployed cloud routes; its playground is not a documented operator-funded general developer inference API.

Publishes open models and resources, but no stable public developer inference endpoint with current auth and quotas.

Current models have positive prices and credits must be prepaid; Start Free is signup wording, not a compute allowance.

Slots are full and the agreement ended 2026-04-30; historical allowance was 100 accelerator-card hours, 500 GB, and 10 million model tokens.

No current provider-funded inference offer was verified.

Current Starter plans begin at $19/month with shared model/agent/compute credits. Public model integrations do not establish an automatic free direct API allowance; retain exclusion unless a current provider-funded quota is documented.

Non-SwissAI use consumes project credits converted from allocated node-hours at a published CHF rate.

The former public alpha advertised positive model token prices. The current site redirects to Darkbloom's solar-compute product and no current documented free developer inference offer or pricing catalog was found.

The current public catalog has positive per-token or per-execution prices and no permanent recurring free quota.

A separate prepaid inference balance is mandatory; the free router preview does not make routed model inference free.

The live production service documents positive per-token prices, prepaid balance top-ups, and a USD 5 minimum top-up; no automatic free API allowance was found. A separate limited free web chat does not qualify as API inference.

Application and allocation infrastructure is not an automatic shared developer inference entitlement.

The Builder Grant URL now renders the Startup Program ($1,000 approved entry, growth to $5,000 tied to paid usage), not the former $25/$100/$250 packs. Track the current program under fal_ai_startup_program instead.

Sandbox coupons cannot fund API or Workflow calls; only the application-based Builder Grant is potentially free.

A public chat and a first-party/partner case study claim an API platform, but no stable developer base URL, authentication flow, free limits, or API reference could be verified.

The lowest current plan is paid and monthly credits belong to paid subscriptions.

Current billing docs say new accounts receive no automatic credit; billing address and card are required for an API key and at least $5 must be added before ordinary inference. The Free label is a rate-limit tier, not zero-price inference.

The CLI/router aggregates upstream services such as NVIDIA and OpenRouter using user-supplied credentials; it does not fund hosted inference.

Users self-host the router and supply upstream provider keys.

Current docs describe promotional credits from promo codes or discretionary promotions but promise no automatic signup amount; ordinary access requires purchased credit with a $10 minimum.

No current provider-funded inference offer was verified.

Many Spaces expose callable Gradio endpoints, but there is no shared quota, catalog guarantee, authentication policy, or availability commitment; individual Spaces may sleep or disappear.

No current provider-funded inference offer was verified.

The free plan covers gateway requests and tracing; all live model entries have positive inference prices.

The $0 plan explicitly has no API access; API access starts on a paid plan.

Research allocations exist, but the July 2026 update says the shared inference service is still being prepared.

Current AI Studio requires purchased credits and has no verified automatic free allowance.

Lambda sunset the hosted API after 2025-09-25; the current alternative is paid GPU infrastructure.

Public chat and downloadable open weights exist, but no documented general developer inference API was found.

Current free offer covers downloadable/on-device models and tooling, not operator-funded remote inference.

Publishes weights, corpora, and tooling rather than a stable hosted inference API.

The $0 gateway subscription still requires account top-up and charges model list prices; the gateway fee alone is free.

No current provider-funded inference offer was verified.

Current docs and terms describe prepaid positive-price API usage with no quantified automatic free allowance.

The .com service is a downloadable/configurable BYOK gateway and is unrelated to the possibly qualifying hosted nexusrouter.net record.

Current onboarding requires adding at least $5 credit. A statement that early users may receive unspecified promotional offers is neither automatic nor a quantified current free allowance.

Hosted inference uses x402 or OPG payment rather than a free tier.

Free compute and model-development resources exist, but no automatic shared developer inference entitlement is established. A SpeedRun contest offers first-place winners an optional core-model API pass for at most 30 days; exact models, API base/auth and quantitative limits are unpublished.

A previous live call returned MissingBlocksError because no peers held the required model blocks. The current repository still describes an interruptible research/development community API, not guaranteed hosted capacity; this audit did not repeat generation.

Starter credit followed by positive per-token wallet billing.

No current provider-funded inference offer was verified.

Normal inference requires a positive balance; selective startup credits are recorded separately as conditional.

The model weights may be free, but remote hosted compute is not being provided.

The public alpha exposes localhost:8800 after users install a daemon and depends on peers; it still has no stable operator-hosted public gateway or funded catalog.

Tencent's May 12 retirement notice stopped new purchases and API-key creation on June 30 and shut the legacy Hunyuan console and APIs on September 30, 2026. The retained 1M generation and 1M embedding one-year packages are historical, not currently callable.

Official billing docs say there are no free trials; API access requires at least a $5 credit purchase.

The direct Jev API is waitlist-gated and publishes positive token pricing with no automatic free allocation. OpenCode Zen's temporary jev-1.13-free route is recorded under OpenCode rather than as a TypeSafe direct free offer.

No current provider-funded inference offer was verified.

Current terms describe paid Wafer Pass and inference; third-party claims of a free route lack first-party support.

First-party pages still advertise a completely free community pool and document api.webinfer.dev, but that documented hostname again returned no DNS records on 2026-10-06. Callability, catalog, signup, and limits remain unverified; alternative gateways were not re-tested.