---
title: "Unbiased Pareto free inference"
description: "A finite allowance for new accounts; it does not recur."
canonical_url: "https://freeinferencing.com/provider/unbiased_pareto/"
md_url: "https://freeinferencing.com/provider/unbiased_pareto.md"
last_updated: "2026-09-19"
---

# Unbiased Pareto

> 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


## Models mentioned

- `pareto-26.9`

## Limits and terms

```yaml
duration_days: 7
free_evaluation_allocation: amount_not_published
rate_limit: unpublished
recurrence: false
availability: Anyone may request an evaluation allocation; new beta accounts are
  manually reviewed.
```

## What happens to your prompts?

**Partially private.** Unbiased says content is not retained beyond processing or used for training without written consent, but its newer data-policy notice says exact prompt and response retention terms are supplied before onboarding rather than publicly.

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: partial
  plan_scope: Unbiased Pareto seven-day free evaluation allocation and beta API.
  prompt_retention: General terms say submitted content is not retained beyond
    processing except when required by law or for abuse prevention for up to 30
    days, while the newer data-policy notice says the exact raw prompt retention
    window is supplied to members before onboarding.
  response_retention: The same public ambiguity applies to model responses; the
    member-specific onboarding terms govern the exact window.
  ordinary_logging: Account, request, usage, security, and abuse-prevention
    metadata are processed; the public terms do not provide a complete
    field-level retention schedule.
  model_training: Unbiased says it does not use submitted content to train or
    improve models without written consent.
  product_improvement: Content-based improvement requires written consent under
    the general terms; aggregate operational analysis may still occur.
  human_or_operator_access: Authorized personnel may access content when necessary
    for service operations, support, security, abuse review, or legal
    compliance.
  subprocessors_and_routing: Requests may be routed through disclosed
    subprocessors and model providers including AWS, Anthropic, OpenAI, and xAI,
    subject to the selected route and onboarding terms.
  deletion_controls: Account or legal requests may apply, but no public
    self-service per-request deletion control is documented.
  caveat: Strong no-training language is favorable, but the public policy defers
    the controlling prompt and response retention window to member-specific
    pre-onboarding terms.
agreements:
  terms_and_data_policy: https://unbiased.ai/terms/
```

## Eligibility

```yaml
account_required: true
application_or_manual_review_required: true
payment_method_required: not_explicitly_documented
```

## API compatibility and modalities

- **Compatibility:** OpenAI Compatible
- **Modalities:** Model inference


## Before you build with Unbiased Pareto

### 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 1 model ID. 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 **Partially private** because Unbiased says content is not retained beyond processing or used for training without written consent, but its newer data-policy notice says exact prompt and response retention terms are supplied before onboarding rather than publicly. 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 4 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

- [Pricing and free evaluation allocation](https://unbiased.ai/pricing/): Official Pricing
- [Unbiased platform](https://platform.unbiased.ai/): Official Product
- [Terms and data policy](https://unbiased.ai/terms/): Official Terms
- [Pareto evals](https://github.com/circuitandchisel/pareto-evals): Official Repository

## Sitemap

See the full [semantic sitemap](/sitemap.md) for every page and markdown mirror.