Code mode can help smaller LLM models
Researchers have found that smaller LLM models can be improved by using a code mode to call tools, rather than relying on the model to generate a signature of the tool call. This approach has been shown to be effective, especially when multiple tool calls are involved, and can help reduce token spends. The code mode concept was likely introduced by Cloudfare, and is now being used by companies such as Anthropic to improve the performance of their LLM models.
- ▪Smaller LLM models can struggle to generate accurate tool call signatures, which can limit their effectiveness.
- ▪Using a code mode to call tools can help improve the performance of smaller LLM models, especially when multiple tool calls are involved.
- ▪The code mode concept has been adopted by companies such as Anthropic, which uses it to reduce token spends and improve the performance of their LLM models.
Hacker News (AI / LLM) files mainly under ai. We currently carry 3,300 of its stories.
Story provenance
Source · retrieval · rights · ranking — open for full record
inspect →
Story provenance
Attribution is not the same as permission. This drawer separates discovery metadata, excerpts, WeSearch-generated summaries, reuse status, and whether the publisher receives the visit. Nothing here claims a legal grant the publisher has not made.
Record
| Original publisher | InstaVM |
| Canonical URL | https://instavm.io/blog/how-code-mode-can-help-smaller-llm-models |
| Publication time | Sat, 25 Jul 2026 08:46:06 +0000 |
| Retrieval time | 2026-07-25T08:52:10.762Z |
| Last seen | 2026-07-25T08:52:10.762Z |
| Headline source | Publisher (no WeSearch rewrite) |
| Excerpt source | publisher body |
| Excerpt method | First ~120 words (~800 chars) of extracted publisher body, fair-use limited. |
| Summary | WeSearch · cerebras-chat (WeSearch summarizer) |
| Summary source text | contentText |
| Citation coverage | Summary is a WeSearch-generated derivative; primary citation is the original publisher URL. |
| Cluster | 2C3vEy4tLhes |
| Cluster logic | Grouped by semantic title/content similarity across sources within a rolling window. Same-publisher template collisions are excluded from coverage comparison. |
| Ranking reason | Story pages are not engagement-ranked. Hub feeds use recency, with optional source-diversified chronological ordering (cap consecutive stories per source). No personalized ranking. |
| Publisher visit | Yes — open original |
| Substitutes article? | No — link-out required for full text |
Rights status (four layers)
WeSearch handling by dimension
| Indexing | May the item be indexed (stored, ranked, made findable)? | Allowed |
| Snippet | May a short excerpt of the publisher's text be shown? | Allowed |
| AI summary | May WeSearch generate its own short summary of the article? | Limited |
| Retrieval / RAG | May the content be exposed for third-party retrieval-augmented generation? | Not asserted |
| Model training | May the content be used to train AI models? | Not asserted |
| Commercial reuse | May the content be reused commercially? | Not permitted |
Basis: Derived from the published RSS/Atom feed. Contact: [email protected]. Reviewed: 2026-07-24.
Opening excerpt (first ~120 words) tap to expand
It's been clear that LLMs alone are not as powerful as "LLMs with tools". How does LLM call tools? It can generate a signature of the tool call, somewhat like a function call search_web("query") or bash("ls -l") etc. This usually calls a function written to either call the google API, exa, or execute a bash command locally. This is kind of fine as long as tool calls are properly formatted — some slightly less powerful models often make mistakes in generating those signatures. The training was so tedious that lot of models were released without the "tool calling" ability. We had to include training data involving special tokens, for example - |tool_call|.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at InstaVM.
Discussion
0 commentsMore from InstaVM
