37 posts tagged with "AI Engineering"

Amazon Nova is easy to underrate if you only compare it on public leaderboards against GPT-5.6 or Claude 5 — it isn't trying to win that fight. Nova's actual pitch is that if you're already an AWS shop, the model, the customisation tooling, the agent framework and the deployment surface are the same product, with the same IAM, the same billing, and the same on-call rotation. That's a genuinely different value proposition, and it's the one you should evaluate honestly rather than dismissing Nova as "the AWS also-ran."

Anthropic's Claude 5 launch is really three separate launches with one shared story: capability is moving down the price ladder faster than most teams' architecture assumes. Opus 5 landed on 24 July 2026 at the exact same price as the outgoing Opus 4.8 — $5 input / $25 output per million tokens — while offering Fable-adjacent performance on several benchmarks. If your Claude cost model still assumes "the good model is expensive," it's already out of date.

Claude Code is easy to misread as "chat that writes files." It isn't. It is an AI coding agent with an agentic loop, a context window, tools, and a permission system — you steer it through Explore → Plan → Code → Commit, then harden the workflow with CLAUDE.md, skills, subagents, MCP, and hooks.

"Sovereign AI" gets used loosely enough in vendor marketing that it's worth being precise about what Cohere's Command A+ actually offers versus what the phrase implies. Command A+ (218B and 25B, Apache 2.0) is built specifically for enterprises and governments that need to run a capable model entirely within their own infrastructure boundary — not as a philosophical statement, but as a concrete set of deployment, data-residency, and control guarantees that a hosted API fundamentally cannot provide, regardless of that API vendor's compliance certifications.

DeepSeek's real contribution to the 2026 model landscape isn't any single model — it's three specific technical ideas that the rest of the industry has since absorbed to varying degrees: Multi-head Latent Attention for cheaper KV-cache memory, Mixture-of-Experts done at extreme sparsity, and Group Relative Policy Optimisation as a reinforcement-learning method that made reasoning-focused training dramatically cheaper. If you understand those three ideas, you understand why DeepSeek mattered, and you'll recognise the same ideas (in variant form) inside half the other models in this series.

Artificial intelligence models are improving at a remarkable pace. A model released only six months ago can quickly appear less capable, less efficient and less reliable than a newer generation. It is tempting to assume that every new model is smarter because it contains more parameters — but modern frontier systems such as Claude Fable 5, Claude Opus 5, GPT-5.6, Gemini 3.6 Flash and Kimi K3 are improving through a much broader stack of advances.

Most models in this series added agentic and coding capability on top of a general-purpose foundation. GLM-5 is positioned the other way around: an architecture and training programme explicitly aimed at agentic software engineering — multi-step tool use, long-running autonomous tasks, and code-heavy reasoning — as the primary target, with general chat capability as a secondary consequence rather than the main event. That inversion of priorities is the thing actually worth evaluating here, not a benchmark score.

Google's Gemini strategy in 2026 is breadth over singular flagship supremacy: text, images, audio, video, real-time voice, search grounding and Workspace integration under one API surface, refreshed on a fast Flash cadence while the headline Pro upgrade quietly slips month after month. That's not a criticism — it's a fact about how to plan around this family. If your architecture assumes "we'll upgrade to Gemini 3.5 Pro when it ships," you need a fallback plan, because as of this writing it still hasn't.

Gemma 4 is the model family to reach for when the question isn't "which frontier model is smartest" but "which model can I actually run on the hardware I already have." Spanning from a 2B-effective-parameter model that runs on a laptop to a 31B model that needs a real GPU, Apache 2.0-licensed and multimodal out of the box, Gemma 4 is Google's answer to on-device and edge deployment — a genuinely different design target from Gemini's cloud-platform breadth.

Kimi K3 is the model that forces a lot of "just use the closed frontier API" arguments to actually justify themselves with numbers. It's a 2.8-trillion-parameter Mixture-of-Experts model with only 104B parameters active per token, a genuine 1M-token context window, and weights you can download and run on your own hardware today. It is also not, strictly, open source — and that distinction matters more than most teams evaluating it realise.