--- title: "Public AI Inference Utility free inference" description: "A finite allowance for new accounts; it does not recur." canonical_url: "https://freeinferencing.com/provider/public_ai/" md_url: "https://freeinferencing.com/provider/public_ai.md" last_updated: "2026-09-19" --- # Public AI Inference Utility > A finite allowance for new accounts; it does not recur. ## Classification - **Directory:** Current - **Free access:** One-time trial - **Status:** Current trial - **Confidence:** High - **Payment card:** No - **Account:** Required - **Equivalent paid value:** Not quantifiable - **API endpoint:** `https://api.publicai.co/v1` ## Models mentioned - `swiss-ai/apertus-v1.5-8b` - `swiss-ai/apertus-v1.5-8b-thinking` - `swiss-ai/apertus-v1.5-70b` - `swiss-ai/apertus-v1.5-70b-thinking` - `swiss-ai/apertus-8b-instruct` - `swiss-ai/apertus-70b-instruct` - `aisingapore/Gemma-SEA-LION-v4-27B-IT` - `aisingapore/Qwen-SEA-LION-v4-32B-IT` - `speakleash/Bielik-11B-v3.0-Instruct` ## Limits and terms ```yaml free_tier_requests_per_minute: 100 starter_credit_amount: not_documented overage: Positive token prices are deducted from the wallet after starter credit. ``` ## What happens to your prompts? **Not private.** Default terms permit prompt/output retention, service improvement, external routing, and disclosure of raw database material to vetted academic researchers unless the user opts out. Privacy is audited separately from price. Review the current governing terms before sending sensitive or regulated data. ## Data governance and agreements ```yaml data_governance: review_status: reviewed plan_scope: Public AI Inference Utility starter-credit API, account chat, and uploaded content under the current terms. prompt_retention: Public AI stores prompts and uploaded content; chats are retained for 30 days and backups can remain for 90 days. response_retention: Outputs are stored with chats for 30 days and can remain in backups for 90 days. ordinary_logging: Service logs are retained for 90 days; account, identity, usage, and wallet metadata are collected separately. model_training: Contribution to Public AI's model-data flywheel is described as a separate opt-in, but routed model providers can have their own terms and no end-to-end no-training commitment applies to every route. product_improvement: Users grant Public AI rights to use content to operate, analyze, and improve the service. human_or_operator_access: Raw database material can include identity and content and may be shared with vetted academic researchers by default unless the user opts out. subprocessors_and_routing: API requests can route to external inference providers that process plaintext content under their own policies. deletion_controls: Users can opt out of academic-research sharing and request deletion, subject to the 30-day chat, 90-day log, and 90-day backup windows and legal exceptions. caveat: The terms do not promise encryption at rest. Default content storage, improvement rights, external routing, and academic-research disclosure make this unsuitable for sensitive prompts. agreements: terms_and_conditions: https://publicai.co/tc ``` ## Eligibility ```yaml account_required: true payment_method_required: not_documented ``` ## API compatibility and modalities - **Compatibility:** OpenAI Compatible, Huggingface Inference Provider - **Modalities:** Model inference ## Before you build with Public AI Inference Utility ### Read the classification narrowly This record 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 account and region. “One-time trial” should be interpreted together with the Current directory placement, High confidence, No payment-card status, and the plan-specific sources below. ### Resolve the live model route The snapshot records 9 model IDs. Match the exact ID against the provider's current catalog before using it in code because zero-price routes, aliases, context limits, and feature support can rotate while an older documentation page remains online. ### Match the quota to the workload shape Translate the allowance into peak requests per minute, input and output tokens, concurrency, retries, and every daily or monthly ceiling that applies to the intended account. The first limit reached by the workload is the practical ceiling. A large token pool can still fail interactive bursts, and a high request limit can still fail long-context work. Include tool calls and retry traffic, test the largest realistic payload, and treat research, experimental, and community access as conditional on their eligibility and fair-use terms. ### Preserve billing and privacy boundaries The recorded payment-card requirement is No, and account access is required. Confirm both in the live signup flow, set provider-side budgets when charges are possible, and observe the balance or usage fields after a complete request. The prompt-handling classification is **Not private** because Default terms permit prompt/output retention, service improvement, external routing, and disclosure of raw database material to vetted academic researchers unless the user opts out. Re-read the terms for the exact route and plan before sending sensitive, regulated, or proprietary content. ### Follow the evidence, then re-check it This record links 8 first-party sources covering the offer, catalog, limits, pricing, terms, privacy, or adoption evidence available to the audit. Prefer the newest and most specific governing document or live catalog when sources disagree. A dated finding can become stale even when the page remains online, so retain the source and snapshot that supported the decision and submit a correction when a provider changes a material term. ### Plan fallback without policy drift A fallback should preserve modality, context length, streaming, structured output, tools, safety controls, and data terms, not only API syntax. Decide which errors may retry, cap retry storms, and prevent an exhausted free route from silently switching to a billable model. Rotating aliases and free pools can change behavior without changing the endpoint, so retain the response model field and test at least one substitute route before the primary offer becomes unavailable. ### Monitor the offer as a dependency Capture the model ID, response model, rate-limit headers, usage fields, latency, HTTP status, and provider request identifier. Watch authorization failure, quota exhaustion, catalog removal, policy revision, and balance movement as separate failure modes. Re-check the live catalog and governing sources on a schedule proportionate to the workload's importance, and keep an owner and exit path for any production dependency on volatile free capacity. ### Test one complete request before scaling Start with the smallest permitted request using the exact credential, model ID, endpoint, region, and account type intended for deployment. Record the status, headers, usage fields, response model, latency, and dashboard balance movement. Then exercise an invalid model, quota exhaustion, or rate limit so failure is explicit and cannot silently switch to a paid route. Validate streaming, structured output, and tool calls separately because a free model can expose fewer features than its paid counterpart. Keep a budget ceiling outside the application whenever billing is possible, and do not send sensitive data until the observed route matches the reviewed agreement. ## Primary sources - [Public AI API docs](https://platform.publicai.co/docs): Official Docs - [Plans](https://platform.publicai.co/plans): Official Pricing - [Models](https://platform.publicai.co/models): Official Catalog - [Public AI joins Inference Providers](https://huggingface.co/blog/inference-providers-publicai): Historical Announcement - [Public AI Terms and Conditions](https://publicai.co/tc): Official Terms - [Public AI Inference Utility GitHub organization](https://github.com/forpublicai): Code Hosting Stats - [Public AI Inference Utility organization on Hugging Face](https://huggingface.co/publicai): Official Repository - [Public AI listed as a Hugging Face inference provider](https://huggingface.co/docs/inference-providers/providers/publicai): Partner Page ## Sitemap See the full [semantic sitemap](/sitemap.md) for every page and markdown mirror.