Benchmarks — codedb

Query latency

codedb repo (20 files, 12.6k lines)

Query	codedb MCP	codedb CLI	ast-grep	ripgrep	grep	MCP speedup
File tree	0.04 ms	52.9 ms	·	·	·	1,253x vs CLI
Symbol search (`init`)	0.10 ms	54.1 ms	3.2 ms	6.3 ms	6.5 ms	549x vs CLI
Full-text search (`allocator`)	0.05 ms	60.7 ms	3.2 ms	5.3 ms	6.6 ms	1,340x vs CLI
Word index (`self`)	0.04 ms	59.7 ms	n/a	7.2 ms	6.5 ms	1,404x vs CLI
Structural outline	0.05 ms	53.5 ms	3.1 ms	·	2.4 ms	1,143x vs CLI
Dependency graph	0.05 ms	2.2 ms	n/a	n/a	n/a	45x vs CLI

merjs repo (100 files, 17.3k lines)

Query	codedb MCP	codedb CLI	ast-grep	ripgrep	grep	MCP speedup
File tree	0.05 ms	54.0 ms	·	·	·	1,173x vs CLI
Symbol search (`init`)	0.07 ms	54.4 ms	3.4 ms	6.3 ms	3.6 ms	758x vs CLI
Full-text search (`allocator`)	0.03 ms	54.1 ms	2.9 ms	5.1 ms	3.7 ms	1,554x vs CLI
Word index (`self`)	0.04 ms	54.7 ms	n/a	6.3 ms	4.2 ms	1,518x vs CLI
Structural outline	0.04 ms	54.9 ms	3.4 ms	·	2.5 ms	1,243x vs CLI
Dependency graph	0.05 ms	1.9 ms	n/a	n/a	n/a	41x vs CLI

Token efficiency

Structured results, not raw dumps

codedb returns structured, relevant results. For AI agents, this means dramatically fewer tokens per query compared to grep/ripgrep raw output.

Repo	codedb MCP	ripgrep / grep	Reduction
codedb (search `allocator`)	~20 tokens	~32,564 tokens	1,628x fewer
merjs (search `allocator`)	~20 tokens	~4,007 tokens	200x fewer

Indexing speed

Cold start to ready

codedb builds all indexes on startup: outlines, trigram, word, dependency graph. After startup, the file watcher keeps indexes updated. Single-file re-index: <2ms.

Repo	Files	Lines	Cold start	Per file
codedb	20	12.6k	17 ms	0.85 ms
merjs	100	17.3k	16 ms	0.16 ms
openclaw/openclaw	11,281	2.29M	75 s	6.66 ms
vitessio/vitess	5,028	2.18M	50 s	9.95 ms

Query Latency (ms, log scale)

Token Efficiency (search 'allocator')

Indexing Speed by Repo Size

MCP Speedup vs CLI

Feature	codedb MCP	codedb CLI	ast-grep	ripgrep	grep	ctags
Structural parsing	Yes	Yes	Yes	No	No	Yes
Trigram search index	Yes	Yes	No	No	No	No
Inverted word index	Yes	Yes	No	No	No	No
Dependency graph	Yes	Yes	No	No	No	No
Version tracking	Yes	Yes	No	No	No	No
Multi-agent locking	Yes	Yes	No	No	No	No
Pre-indexed (warm)	Yes	No	No	No	No	No
MCP protocol	Yes	No	No	No	No	No
Full-text search	Yes	Yes	Yes	Yes	Yes	No
Atomic file edits	Yes	Yes	Yes	No	No	No
File watcher	Yes	Yes	No	No	No	No

Architecture

Why codedb is fast

MCP server

Indexes once on startup. All queries hit in-memory data structures with O(1) hash lookups. No filesystem access at query time.

CLI tools

~55ms process startup + full filesystem scan on every invocation. Even simple queries pay the full cost.

ast-grep

Re-parses all files through tree-sitter on every call. ~3ms for small repos, scales linearly with codebase size.

ripgrep / grep

Brute-force scans every file on every call. ~5-7ms for small repos. No structural understanding, no index.