Route coding-agent requests across models
Weave Router 2.0 routes coding-agent requests to models based on available quota and task needs.
Coding / product dossier
Weave Router 2.0
Routes coding-agent requests across Claude and GPT plans by quota, task complexity, and cost.
Product brief
What Weave Router 2.0 does.
Use Claude models in Codex and GPT models in Claude Code on your existing plans, routed to whichever has quota left. Weave Router 2.0 routes each coding agent request to the cheapest model that can get it right. On Terminal-Bench 4.0 and SWE-Atlas it matches GPT-6 Astra at half the cost, 2x faster. Powered by a new classifier that scores task complexity and cache-aware switching that only moves when savings beat the rebuild cost.
Why we selected it
Addresses a concrete coding-agent cost and quota problem with task-complexity classification and cache-aware model switching. It also naturally fits the current focus on making coding-agent workflows more inspectable and
- Best for
- Developer tools
- Category
- Coding
- Daily picks
- 1
- First selected
- 2026-09-16
Product preview saved with our daily selection
Capability scan
What it can help with.
Only capabilities supported by the product information we collected are listed here.
Use models across existing plans
It is described as enabling Claude models in Codex and GPT models in Claude Code on existing plans.
Classify task complexity
A classifier scores task complexity to inform routing decisions.
Apply cache-aware switching
The router switches only when the stated savings exceed the cost of rebuilding cache.
Select for stated cost efficiency
It is designed to route a request to the cheapest model that can get it right.
Best-fit use cases
Continue coding-agent work when one model plan has no remaining quota.
Route tasks between Claude and GPT models in Codex or Claude Code.
Use task-complexity classification to choose a model for a coding request.
Avoid model switches when cache rebuild costs outweigh projected savings.
FAQ
Before you open it.
What does Weave Router 2.0 route?
It routes coding-agent requests among models.
Which coding environments are named?
The description names Codex and Claude Code.
How does the router decide whether to switch models?
It uses a task-complexity classifier and cache-aware switching, moving only when stated savings exceed the cache rebuild cost.
What quota problem does it address?
It is intended to route requests to whichever named model plan has quota remaining.
What benchmark claim is provided?
The product description says it matches GPT-6 Astra at half the cost and twice the speed on Terminal-Bench 4.0 and SWE-Atlas.