Guides & Explainers
In-depth written guides on AI concepts, from beginner to advanced
Large Language ModelsGPT-6 Astra: A New Generation at 2.5x the Price
GPT-6 Astra is OpenAI’s first sixth-generation model, with a 1,050,000 token context window at $10/$50 per million tokens. What OpenAI publishes, and what it does not.
Claude Fable 5.1: The Price Change Is in the Cache, Not the Token
Claude Fable 5.1 keeps Fable 5’s $10/$50 token pricing but cuts cache reads to a quarter. What that changes, and what it does not.
Qwen3.7 Max: Alibaba's Agent-Centric Flagship
Alibaba positions Qwen3.7 Max around agent-centric use rather than raw chat quality - here's what its 1M context window, pricing, and caching actually cost.
Muse Spark 1.2: Meta's Pivot From Open Weights to a Closed API
Muse Spark 1.2 is a proprietary, API-only model from Meta - a break from the open-weight Llama line. Here's what shipped, what it costs, and how it ranks.
DeepSeek V4-Pro's Peak/Off-Peak Pricing, Explained
DeepSeek V4-Pro halves its price outside 01:00-04:00 and 06:00-10:00 UTC. Here is how that billing model works and how it compares to flat-rate competitors.
Grok 4.5: The Model Grok 4.6 Replaced, Not Retired
xAI's Grok 4.5 is still available and still competitively priced, even after Grok 4.6 shipped on the same foundation. Here's what changed and what didn't.
Claude Sonnet 5: How a "Temporary" Price Became Permanent
Claude Sonnet 5 launched at introductory $2/$10 pricing through August 31, 2026. On August 10, Anthropic made it permanent. Here's what else changed from Sonnet 4.6.
Claude Opus 4.8: What Anthropic's Legacy Flagship Still Offers
Claude Opus 4.8 is the Anthropic model Opus 5 superseded - same $5/$25 pricing, same 1M context window, still generally available. Here is when it still makes sense.
Gemini 3.1 Pro Is Still a Preview Model - Here's What That Means for Pricing
Google lists Gemini 3.1 Pro as a preview model, not stable - and its pricing doubles past 200k input tokens. Here's what to know before building on it.
GPT-5.6's Three-Tier System: Sol, Terra, and Luna Compared
OpenAI ships GPT-5.6 in three pricing tiers - Sol at $5/$30, Terra at $2/$12, and Luna at $0.20/$1.20 per 1M tokens. Here is what actually differs between them.
Claude Fable 5: The Tradeoff Behind Anthropic's #1 Model
Claude Fable 5 ranks #1 on LMArena's text leaderboard and tops Anthropic's own lineup for intelligence. It's also the slowest and most expensive Claude model - here's the actual tradeoff.
Claude Opus 5: Anthropic's Coding Flagship, and When to Still Use Opus 4.8
Claude Opus 5 is Anthropic's model for complex agentic coding, at $5/$25 per 1M tokens with a 1M context window. Here is what changed from Opus 4.8, which is still available at the same price.
Showing 1–12 of 19 guides