Tag feed

#openai

2,093 posts243 followers

Explore Hashnode

Alternatives

Trending tags this week

MMagnusmodelflare.hashnode.dev12h ago · 10 min read

Function Calling: Responses API vs Chat Completions

A wire-level comparison of function definitions, Call IDs, result messages, streaming arguments, authorization, idempotency, and route compatibility. Responses and Chat Completions can both ask an app

0

DHDylan Hogginfocruncher.hashnode.dev1d ago · 9 min read

Why AI automated maths early, and how mathematicians are reacting

How mathematicians responded to OpenAI's October 2026 maths release, and what their responses suggest about the order in which AI automates different kinds of work. 1. Why AI automated maths early In

0

RDRONI DASsystemdesignacademy.hashnode.dev1d ago · 13 min read

GPT-6.1 Sol costs 20 times more than Luna. On my 300 support tickets, they tied.

I sent the same 300 support tickets to two OpenAI models. One costs 20 times more than the other. Both sent 299 of the 300 tickets to the right team. The expensive one was GPT-6.1 Sol. The cheap one w

0

MMihai_LeanZeroleanzero.hashnode.dev1d ago · 33 min read

Claude Haiku 5.5 vs GPT-6 Luna: Cost, Benchmarks

Claude Haiku 5.5 vs GPT-6 Luna is the small-model match-up Anthropic picked for its own launch chart. So we ran both on our two agentic coding benchmarks. Haiku won both. On Forge 1.0, an Atlassian Fo

0

ZZyVOPblog.zyvop.com2d ago · 11 min read

OpenAI’s Math: Fast Generation, Slow Verification

OpenAI's math release has been public since October 6. The first bill for checking it has already arrived. Open github.com/openai/math and the first thing you notice is what's missing: no benchmark ha

0

SMShubham Moretheshubhammore.hashnode.dev3d ago · 12 min read

Building with LLMs? Understand Tokens, API Costs, and How to Avoid Unexpected Bills

Introduction: The Part of LLM Development We Often Overlook When we start building applications with Large Language Models (LLMs), we're usually excited about what we can create. We experiment with di

0

江江流石jliushi.hashnode.dev3d ago · 2 min read

Codex CLI + Custom Relay: The config.toml You Actually Need

The minimum config to run Codex CLI through a third-party relay. Based on the AI Relay setup guide. Relays see your requests; pick one you trust. 1. Pick a relay that supports Codex Most relays on the

0

SKShivam Kumarshivaysinghrajput.hashnode.dev3d ago · 3 min read

AI Watermarking Just Got Its First Industry Standard

AI Watermarking Just Got Its First Industry Standard On October 7, 2026, Google opened its SynthID Detector to the public globally at synthid.com. Anyone can now check images, video, and audio for Syn

0

SRSai Ram Kothasairamkotha.hashnode.dev4d ago · 8 min read

Reasoning Effort Across AI Providers: One Knob, Four Meanings

Every major AI provider now lets you turn reasoning up or down. The names look alike: effort, thinking level, thinking budget. The behavior is not alike. Some settings are a hint, some are a hard cap,

0

RDRONI DASsystemdesignacademy.hashnode.dev4d ago · 11 min read

OpenAI's Decisions API looked at a clean egg and said 47% dirty. Here is what your code should do with that.

On 6 October 2026 OpenAI released the Decisions API in public beta. It does something most AI APIs do not: it does not write a reply. You send it a photo or some text, a question, and the list of answ

0

#openai

Search Hashnode

#openai

Explore Hashnode

Trending tags this week

Function Calling: Responses API vs Chat Completions

Why AI automated maths early, and how mathematicians are reacting

GPT-6.1 Sol costs 20 times more than Luna. On my 300 support tickets, they tied.

Claude Haiku 5.5 vs GPT-6 Luna: Cost, Benchmarks

OpenAI’s Math: Fast Generation, Slow Verification

Building with LLMs? Understand Tokens, API Costs, and How to Avoid Unexpected Bills

Codex CLI + Custom Relay: The config.toml You Actually Need

AI Watermarking Just Got Its First Industry Standard

Reasoning Effort Across AI Providers: One Knob, Four Meanings

OpenAI's Decisions API looked at a clean egg and said 47% dirty. Here is what your code should do with that.