docs: Cloudflare Workers AI now honors the agent reasoning level

Open Beginner friendly
  • Is something wrong?
#138 0 comments 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 1 day

Nobody has claimed this yet.

Assessment

Difficulty
2/5
Estimated time
1-3 hours
Newbie friendliness
78/100
Issue type
Documentation
Clarity
Clearly specified
Activity status
Active
Tech stack
typescript
Domain
documentation

Research direction

Read concepts/providers-and-routing in English and Vietnamese and locate the Reasoning section’s “Each adapter maps the level in its own way” table. Add a Cloudflare Workers AI row describing the model dependent reasoning levels and behavior; consider the optional note in integrations/llm-providers. Done when both language versions document the behavior accurately and the pages build.

Written by the indexing model from the issue text.

Description

ai-handle

What changed for users

Cloudflare Workers AI agents now follow the agent's reasoning setting on the models whose Workers AI schema documents a reasoning effort level:

  • DeepSeek V4.1 Flash 1017, V4 Flash 0731 and V4 Pro 0813: none, low, high, max
  • Kimi K2.6: none, high
  • Qwen 3.8 27B: low, medium, xhigh
  • GLM 5.2: none, high, max
  • GLM 5.3 and GLM 5.3 Flash: low, high, max

On these models the agent settings offer exactly the levels listed above. A level the model does not support is downgraded, as it is on other providers.

off is the default for agents with no reasoning setting. It now turns reasoning off instead of leaving the model at its default effort (high or max), which makes replies faster and uses fewer tokens. GLM 5.3 and 5.3 Flash cannot turn reasoning off, so off runs them at low and hides the reasoning they still produce. auto keeps the model's default. Other Workers AI models are unchanged.

Pages likely affected

  • concepts/providers-and-routing (en + vi): add a Cloudflare Workers AI row to the "Each adapter maps the level in its own way" table in the Reasoning section.
  • integrations/llm-providers (en + vi): optional one-line note that Workers AI reasoning levels depend on the model.

Source

nextlevelbuilder/dewee#2262 (PR linked there).

Dominant language
TypeScript
Stars
2
Forks
2
Avg merge
1d 15h
Merged PRs (30d)
63

Getting set up

This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from nextlevelbuilder/dewee-web-v2

All issues in nextlevelbuilder/dewee-web-v2

Similar issues

More TypeScript issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.