Is CodeAct superseded?

CodeAct (Tool use / function calling): cited as a baseline — critiqued by newer work, not yet beaten on a benchmark here. 1 paper(s) critique it, 0 beat it on benchmarks — #33 of 55 most-superseded. Sub-problem: cluster led by StepTool. Newer alternatives in the same sub-problem include CARL, R2IF.

Tool use / function calling

CodeAct

Superseded baseline#33 of 55 most-superseded

Cited as a baseline — critiqued by newer work, not yet beaten on a benchmark here

1 papers critique it · 0 beat it on benchmarks

What papers say

Verbatim critique sentences, each from a paper that cites CodeAct as a baseline.

relying solely on the built-in functions of programming languages and a limited set of predefined libraries codeact restricts the quantity and scope of tools that can be invoked

— CodeTool: Enhancing Programmatic Tool Invocation of LLMs via Process Supervision

What to use instead

Recent methods in the same sub-problem, not yet superseded in the knowledge base — arXiv benchmark leaders, not vetted production recommendations.

CARLKnowing When to Ask: Segment-Level Credit Assignment for LLM Tool Use
May 27, 2026
R2IFR2IF: Aligning Reasoning with Decisions via Composite Rewards for Interpretable LLM Function Calling
Apr 22, 2026