CheaperInference

CheaperInference is an AI inference discount reseller built and operated by Keak. It buys excess, unused inference capacity commitments from AI companies and resells them through one OpenAI-compatible API, passing most of the discount to customers at up to 30% off list price. The launch date is not publicly disclosed; the service surfaced in August 2026 through a promoted X ad by @fabrice_mayrand that drew about 1.9M views.

Key Features

• One API across providers: Discounted OpenAI, Anthropic, and Google models (gpt-5.4, Claude, and more) behind a single OpenAI-compatible endpoint, with no request-format changes.
• Market-linked savings: Rates follow the excess-capacity market, never above direct list price, with no separate routing surcharge.
• Auto-fallback routing: Requests route in price order; network failures and 404/408/409/425/429/5xx responses retry once, then shift to the next-lowest-cost eligible route.
• Wallet billing: Balance is checked before each request and usage is deducted after completion, with an auto-recharge option.
• Privacy: Prompt and response bodies are not stored in the app database.
• Pricing feed: GET /v1/models exposes pricing_version and pricing_updated_at, with hourly observations and 24h/30d history per model.

Use Cases

• AI startups: Cut inference spend on chatbots, copilots, and RAG pipelines without touching integration code.
• Agent workloads: High-volume loops benefit from price-ordered routing and automatic fallback.
• Teams: Workspaces, team access control, no annual contract, and no setup fee.

Pricing

Billing is usage-based with no monthly commitment and no contract. Fund your wallet from $5 and get a $10 bonus credit after your first successful payment. The all-in rate never exceeds direct list price.

Advantages

Versus direct provider APIs: one key covers many providers, and price-ordered fallback removes custom routing code.

Versus OpenRouter: CheaperInference is a discount reseller built on purchased excess capacity, pitching an at-or-below-list rate rather than marketplace breadth.

Getting Started

1. Sign up at cheaperinference.com.
2. Fund the wallet with $5 or more to claim the $10 first-payment bonus.
3. Create an API key in the dashboard.
4. Change two values: set the base URL to https://api.cheaperinference.com/v1 and swap in your key.
5. Send requests with your existing model, messages, tools, and streaming settings.

FAQ

Do I need to change my request format?

No. Only the base URL and API key change; the rest stays OpenAI-compatible.

What data is stored?

Prompt and response bodies are not stored in the application database.

How are failed requests handled?

Eligible failures are retried once, then routed to the next-lowest-cost route.

Alternatives

• OpenRouter: a broader multi-provider catalog when model variety matters most.
• Direct provider APIs: better for provider-specific features or negotiated committed-use pricing.

Tips

1. Poll the pricing feed to time large batch jobs around pricing_version changes.
2. Enable auto-recharge so requests never fail on an empty wallet.
3. Pilot one workload and compare cost and latency before moving all traffic.

Conclusion

CheaperInference is a focused cost play: if your stack already speaks the OpenAI format, savings of up to 30% are a two-line change away. For price-sensitive teams, it is an easy pilot.