Anthropic Claude Haiku
AssessTools
A smaller, faster Claude model variant optimized for lower-latency and lower-cost tasks.
Why it's here
Placed in Assess: 2 article(s) of evidence from 2 source(s), led by product launches, with 0 in the last 30 days. Confidence 37%.
Evidence (2)
- 3Simon Willison·7/3/2026framework_updateClaude Code advises delegating coding to lower-power subagents
Simon Willison describes a workflow where Claude Code and Fable are instructed to use their own judgment about when to test and when to delegate smaller coding tasks to lower-power models. The goal is to reduce token usage and cost while keeping higher-level judgment in the main model.
- 8The New Stack·7/2/2026product_launchMicrosoft launches $2.5B AI unit focused on model flexibility
Microsoft is creating a $2.5 billion AI adoption unit, Microsoft Frontier Company, to help enterprise customers mix and match AI models instead of relying on a single provider. The move reflects a broader shift toward AI systems that route each request to the model best suited for the task, with Microsoft saying its earlier Copilot approach was too tightly bound to OpenAI models.