Research library
Artificial intelligence
Evaluate where AI is useful, understand the systems behind it, and make informed choices about models, data, and deployment.
194 articles · Page 2 of 17
Claude Haiku 5.5: A 90% List-Price Cut, a 75% Real One, and Five Ways to Get a 400
Haiku 5.5 lists at a tenth of Haiku 4.5's price; Anthropic says about 75% less in practice. The gap, the 400s after switching, and Sonnet 5.5 vs Luna.
Gemini API Preview Shutdowns: Veo 3.1 Moves to Omni, Voice Follows in November
Veo 3.1 and Omni previews may end from October 22, TTS and Live previews from November 17. Cost per clip, voice replacements and the January price rise.
Mistral Large 4: What Is Confirmed Before the Weights Drop, and What Is Not
Mistral Large 4 is a 1T-parameter MoE you can call today at a half-price preview. Weights, license and independent scores are still due. What to check.
Decision Models Go Open Weight: Cloudflare Clef, Liquid d1 and TypeSafe Jev Compared
Cloudflare's Clef and Liquid's d1-3B are open-weight decision models. Compared with Jev and OpenAI's Decisions API on license, price, context and limits.
OpenAI Shuts Down GPT-4, o1 and o4-mini on October 23; Copilot Drops GPT-5.5 on October 19
OpenAI's API removes gpt-4, o1, o3-mini, o4-mini and gpt-image-1 on October 23. Copilot removes GPT-5.5 and GPT-5.4 on October 19. A dated checklist.
OpenAI Decisions API: $0.10 per Million Input Tokens and Nothing for the Answer
OpenAI's Decisions API: public beta, GPT-6 Luna only, $0.10 per million input tokens, no output charge. Where it beats Responses with structured output.
Google Ships EmbeddingGemma 2: Text, Images, Video and Audio in One Vector Space
EmbeddingGemma 2 embeds text, code, image, video and audio in one 768-d space, ungated under Apache 2.0. Real sizes, self-reported scores, what to test.
Nano Banana 2.1: Half the Per-Image Price, Higher Token Rates
Google's Nano Banana 2.1 about halves the per-image price but triples input rates. Nano Banana 2 is deprecated, with no shutdown date. What to check.
Claude Sonnet 5.5: Compare Cost per Finished Task
Sonnet 5.5 keeps Sonnet 5 token prices. Its claimed savings come from using fewer tokens. How to test that claim in your workflow.
Holo4 Combines Computer Use and Tools, with Different Licenses
Holo4 spans GUI, code, MCP and API workflows. Its 27B and 35B-A3B checkpoints have different licenses. What to check before evaluation.
Reflection Beam: From Announcement to Open Weights
Reflection announced Beam on October 5 and released its weights on October 16. A dated look at the 501B model, benchmarks and deployment limits.
OpenAI DevDay 2026: Always-On Dots, GPT-6.1 Sol and Uneven Availability
OpenAI DevDay 2026 announced dots, GPT-6.1 Sol, Ultrafast, Agents API computer use and plugin extensions. What is live now, what it costs and what to test.