Browse SynapseWire coverage, updates, and related reads about AI models.
Topic hubs connect recurring themes across breaking news, hands-on reviews, tutorials, and strategic analysis.
Google priced Gemini 4 Hydrogen at $1.50 and $7.50 per million tokens and opened Claude Meetings in beta on Team and Enterprise Google Meet calls.
Meta priced Muse Ember at $0.10 and $0.50 per million tokens on November 9, with a surcharge after 200,000 tokens and a split between external and internal scores.
GitHub Copilot Fusion reaches GA with Standard, High, and Max effort levels, dynamic multi-model routing, and benchmark parity with Claude Code at lower cost.
GPT-4 pricing dropped from $30 to $1-2 per million tokens. What's driving the collapse, and what does it mean for developers building with AI?
Two frontier AI models launched within two weeks of each other in July 2026. Here's how they actually compare on benchmarks, pricing, and real-world use cases.
The frontier model race is giving way to a more important competition: who can run AI cheaper, who controls the infrastructure, and whether open-source models can match proprietary performance at a fraction of the cost.
Snap is spinning off a generative AI video team into Dotmo, a company building AI models for interactive gaming. How the licence-and-equity deal works, Bobby Murphy's role, and why AI spin-offs are used to manage cost.
A practical guide to the agent-focused features that shipped with Claude Opus 4.6: adaptive thinking, four effort levels, context compaction, 128K output and Claude Code agent teams, with tips on cost and when to use each.
What Anthropic's Claude Opus 4.6 changed: a 1M-token context window in beta, its reported benchmark results, API pricing, new developer controls, and when it is worth using over cheaper models.
An explainer on GLM-4.7-Flash, Z.ai's 30B-A3B MoE open-weight model: published benchmarks against Qwen3-30B and GPT-OSS-20B, local deployment with vLLM, free API access, use cases and limitations.