mediumAI Engineering

How do you choose between open-weight models and proprietary API models?

Q: How do you choose between open-weight models and proprietary API models?

**TL;DR.** **Proprietary APIs** buy top capability with zero infrastructure; **open weights** buy data control, customization and at-scale cost predictability — paid for in serving ops and a capability gap. Decide on **capability need, privacy constraints, volume economics, and team maturity**. proprietary API frontier capability zero infra, pay per token data leaves your VPC* provider controls versions <rect class='d-box-muted' x='240' y='24' widt

469 views

Understand the problem

Build vs buy for models: capability, cost curves, privacy, and operational burden.

open-weightsapibuild-vs-buyarchitecture

Attempt it yourself

Sketch your approach before reading the solution — that's what interviews test.

Stuck? AI Nudge Available

Get a conceptual hint to guide your logic without spoiling the final implementation.

Study the solution

The solution is waiting

Give it an honest attempt first — then compare your thinking with the full walkthrough.

Read the code

Provider-agnostic seam so you can switch

interface LlmClient {
  complete(req: { system: string; messages: Msg[]; maxTokens: number }): Promise<Reply>;
}

// implementations: AnthropicClient, BedrockClient, VllmClient...
// routing by task tier keeps economics honest:
const client = task.tier === "frontier" ? anthropic : selfHostedSmall;

Join the discussion

Discussion (0)

No responses yet. Be the first to share what you think.

What is chain-of-thought prompting and when does it actually help?

How do you reduce hallucinations in a production LLM application?