Answer: Aider uses different "edit formats" to collect code edits from LLMs, including the "whole" format and one or more diff formats. The "whole" format is easiest for an LLM but uses many tokens, while diff formats are more efficient and allow editing larger files. Status: Confirmed Entry: tool/aider/models/benchmark-edit-format Revision: 1 Digest: sha256:02d728464588a288 (changes whenever the content changes) Timeline: Created: 2026-11-13T02:56:22Z Last changed: 2026-11-13T02:56:22Z Last verified: 2026-11-20T11:42:47Z Next check: 2026-11-21T11:42:47Z Check interval: Daily Expected change: No fixed schedule Data: Field | Value | Source | As of edit formats | edit formats | [1] | 2026-11-20 whole format | whole | [1] | 2026-11-20 diff formats | diff formats | [1] | 2026-11-20 default format for lesser known models | whole | [1] | 2026-11-20 Sources: [1] Aider. "aider.chat/docs/leaderboards/notes.html". https://aider.chat/docs/leaderboards/notes.html Role: primary. Accessed: 2026-11-20T11:42:47Z Quote: "Aider uses different “edit formats” to collect code edits from different LLMs." (Supports: edit formats) Quote: "The “whole” format is the easiest for an LLM to use, but it uses a lot of tokens" (Supports: whole format) Quote: "Models which can use one of the diff formats are much more efficient, using far fewer tokens." (Supports: diff formats) Quote: "For lesser known models aider will default to using the “whole” editing format" (Supports: default format for lesser known models) Published: 2026-04-25T16:45:23Z Snapshot SHA-256: 58f709ea4691203fa74781da87ff6bb23a3732e65704746e0d84cccd1c0a0113 (captured 2026-11-15T11:44:26Z) Verification: Method: Automated extraction from [1] with deepseek-flash (JSON output, thinking disabled). Every value was checked by code to appear verbatim in a quoted passage of [1]. On this check the text of [1] was unchanged since 2026-11-13T02:56:22Z, so the values verified then were kept. Cross-check: not-available at 2026-11-20T11:42:47Z. No independent second source is configured for this entry Last successful check: 2026-11-20T11:42:47Z Last attempt: 2026-11-20T11:42:47Z Notes and limitations: Extracted automatically from a single official source (section "Notes on the edit format"). Related: https://latestn.com/tool/aider/models/benchmarking-results-notes What are the notes on benchmarking results for Aider's code editing leaderboard? https://latestn.com/tool/aider/models/best-models-recommendation Which LLMs does Aider recommend as the best models to use? https://latestn.com/tool/aider/models/browser-ui-supported-models Which LLM models work well with Aider's browser UI? https://latestn.com/tool/aider/models/accepts-settings-compatibility Which reasoning models does Aider support and how does accepts_settings work? https://latestn.com/tool/aider/models/anthropic-thinking-tokens How do I enable thinking tokens for Anthropic models in Aider? https://latestn.com/tool/aider/models/bedrock-available-models Which models are available for Aider on Amazon Bedrock? https://latestn.com/tool/ All entries in AI tools Change log: Time | Event | Change | Source 2026-11-13T02:56:22Z | added | Entry created | — Changes for this entry as a feed: https://latestn.com/tool/aider/models/benchmark-edit-format/latest Challenge this entry: Open challenges: None How: If a value here is wrong or out of date, send what is wrong and a link to evidence. Challenges are queued and checked against the evidence; they do not change the entry by themselves. If the entry changes as a result, the change log says so. Form: https://latestn.com/dispute/tool/aider/models/benchmark-edit-format POST: https://latestn.com/dispute/tool/aider/models/benchmark-edit-format with claim, evidence_url, and optionally field and contact (form-encoded or JSON) GET: https://latestn.com/dispute/tool/aider/models/benchmark-edit-format?claim={what is wrong}&evidence_url={link} How to cite: LatestN. (2026). What edit formats does Aider use in its benchmarks? Revision 1, last verified 2026-11-20T11:42:47Z. https://latestn.com/tool/aider/models/benchmark-edit-format (accessed 2026-11-21T09:53:33Z). @misc{LatestN_tool_aider_models_benchmark_edit_format_r1, title = {What edit formats does Aider use in its benchmarks?}, author = {{LatestN}}, year = {2026}, howpublished = {\url{https://latestn.com/tool/aider/models/benchmark-edit-format}}, note = {Revision 1; digest sha256:02d728464588a288; last verified 2026-11-20T11:42:47Z; accessed 2026-11-21T09:53:33Z} } Reuse: License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) Attribution: Copying and adapting are allowed, including commercially, if you credit LatestN and link to the entry you used. Each entry ends with a ready-made citation. URL: https://latestn.com/tool/aider/models/benchmark-edit-format