Ask HN: How to use agents via API cheaply?

3•thih9•1d ago

I’d like to code with an AI agent via a BYOK open source client. I want to be able to inspect everything, check the system prompt, etc. Also, I’d like to avoid vendor lock in.

Is anyone here using a setup like this?

Does this get close in pricing and efficiency to the subsidized plans offered by model providers (like Claude Pro, etc)?

In particular Deepseek and Kimi models are being praised as both efficient and cheap; I’m curious if anyone built a coding setup around them.

Comments

ryanxsim•1d ago

Outside claude and codex, I use kilo code and its good so far

thih9•1d ago

What model do you use with kilo code? What are the typical usage costs?

How would you compare it to the provider plans (in costs and efficiency)?

ryanxsim•1d ago

Any local llm or third party outside gpt and claude. It depends on the trends sometimes deepseek, kimi or glm or sometimes gemini 3.1 pro was decent when it was decent

Ask HN: What's the best hands-on path to learn ML inference infrastructure?

Ask HN: What are the most promising RL fields for a new master student?

BSD seq(1) on Linux

Robotic Adaptation Suffers from Data Collection Bottleneck

Ask HN: Anyone suffering from the tyranny of the Waiting Equation?

Ask HN: Is an in-page speed reader useful?

Ask HN: I spent personal money on game dev. How to turn stalled project around?

Ask HN: Are you using Rust on embedded devices yet? If not, why?

Ask HN: Graal VM Relevance

Ask HN: Something fishy is going on with Google feed on my Android

Ask HN: Protecting your Sites/Services from Unwanted Traffic?

Ask HN: What livestream do you keep open in a tab?

Searchable, leveled learning map of AI/ML tools

Ask HN: Anyone else experiencing extreme Codex limits token burn rate?

Connection-Agnostic Presence Tracking for Stateless Distributed Back Ends

ASK HN: Why has technology become so unreliable?

I built a jav data API, need honest feedback

Ask HN: Which Jabber clients support SCRAM+ and XEP-0474

YouTube Comments in .csv

Ask HN: Is Your Work Relaxing?

Timed YouTube video transcripts in .txt and .pdf

What happens behind the scenes when we change effort for same LLM models?

Ask HN: HotPin – lossless 120B MoE inference on 24GB RAM (CPU, 50 loc)

Tell HN: Namecheap gave my account to an unverified third party

Who's making money from using/exploiting AI?

Telemetry.sh – Telemetry Built for Agents

Ask HN: What do you do when your AI agents are working?

Ask HN: How to get your open-source project gain popularity?

Ask HN: What happens when we do compress the context in Claude Code?

Ask HN: Is neuromorphic computing going to replace traditional AI?