Blog

What changed in AI APIs, and how to fix it.

Model shutdowns, breaking API changes and price moves across OpenAI, Anthropic, Google, DeepSeek and xAI. Every price and date is checked against the provider's own pages.

31 Oct 2026Claude alternatives for coding: GPT-6 Sol, Gemini 3.8 Flash, Kimi K3, DeepSeek and Grok ranked by cost per passing taskGPT-6 Sol, Gemini 3.8 Flash, Kimi K3, DeepSeek and Grok against Claude, ranked by what a coding task that passes its tests costs, plus a Python router.
30 Oct 2026The cheapest way to use the OpenAI API in 2026Current GPT-5.6, GPT-6 and GPT-6.1 prices, and what really cuts an OpenAI bill: a smaller model, prompt caching, Batch or Flex, with a worked monthly example.
29 Oct 2026How much does an LLM API cost per month? Five worked examples at October 2026 pricesMonthly LLM API cost for a support chatbot, RAG search, coding agent, transcription and image app at October 2026 list prices, with the math and a calculator.
28 Oct 2026gpt-4o, gpt-4o-mini and gpt-4.1 shut down on 14 January 2027: what breaks, what gpt-6-luna costs, and a test planOpenAI shuts down gpt-4o, gpt-4o-mini, gpt-4.1 and gpt-4.1-mini on 14 January 2027. What breaks on gpt-6-luna, what it costs, and a test plan to finish first.
27 Oct 2026Gemini 2.5 after the shutdown: "not found", "no longer available to new users", and what to use insteadGemini 2.5 Pro and Flash return 404 for new users, and Vertex AI retired them on 20 October. What the errors mean, what to switch to, and what breaks next.
26 Oct 2026OpenAI "model not found" after the 23 October shutdowns: what replaced o4-mini, gpt-4.1-nano and the restOpenAI shut down o4-mini, gpt-4.1-nano, gpt-3.5-turbo, gpt-4, o1 and more on 23 October 2026. The replacement for each, what it costs, and the fix.
25 Oct 2026DeepSeek V4 Pro vs V4.1 Flash: what changed on 10 September and 17 OctoberDeepSeek retired V4 Flash and now serves V4.1-Flash-1017. How it compares with V4 Pro, what each costs at peak and off-peak, and the peak windows in US time.
24 Oct 2026Claude "temperature is deprecated for this model": what changed on 21 October and how to fix itAnthropic's API version 2026-10-21 rejects temperature, top_p and top_k with a 400. Which models and versions are affected, and the code changes to fix it.
23 Oct 2026Claude API credits: how buying them works, and what to do when the payment failsHow Claude API credits work in the Anthropic Console, the error you get at a zero balance, why card payments fail, how to fix them, and other ways to pay.
22 Oct 2026OpenAI API key spending limits: why you can't cap a single key, and how to get a hard cap anywayOpenAI caps spend per project or organization, never per API key. What each limit blocks, why spend runs past it, and how to give each key a hard cap.
21 Oct 2026The cheapest AI API aggregator in 2026: what $1 of usage really costs on OpenRouter, Requesty, Vercel, Cloudflare and AightPurchase fees, minimums, expiry and refunds compared for OpenRouter, Requesty, Vercel, Cloudflare, Eden AI and Aight, as the real cost of $1 of model usage.
20 Oct 2026One API key for every major LLM: Claude, GPT, Gemini, DeepSeek and GrokCall 126 models from 16 makers with one OpenAI-compatible key and one prepaid balance. Setup, prices by model, per-key caps and 10% extra on every recharge.