README.md

Benchmarks

Quality and speed benchmarks for semble.

• Main results
• Token efficiency
• By language
• Ablations
• Dataset
• Methods
• Excluded methods
• Running the benchmarks

Main results

Quality and speed across all methods.

Method	NDCG@10	Index	Query p50
CodeRankEmbed Hybrid	0.862	57 s	16 ms
semble	0.854	263 ms	1.5 ms
CodeRankEmbed	0.765	57 s	16 ms
ColGREP	0.693	5.8 s	124 ms
BM25	0.673	263 ms	0.02 ms
grepai	0.561	35 s	48 ms
probe	0.387	—	207 ms
ripgrep	0.126	—	12 ms

Time to first result (index + query) vs NDCG@10	Query latency on a warm index vs NDCG@10

The 137M-param CodeRankEmbed Hybrid wins NDCG@10 by 0.008. semble wins index time by 218x and query latency by 11x.

NDCG@10 is averaged across all queries. Speed numbers use one repo per language, CPU only: cold-start index time and warm query p50 (median across 5 consecutive runs).

Token efficiency

Coding agents (Claude Code, OpenCode, etc.) typically find code by running grep on keywords and reading the matched files. We model that workflow and compare it against semble's chunk retrieval across our full benchmark of 1251 queries.

Expected tokens per query

For each query: tokens consumed at first relevant hit, or 32k if the method never finds anything. Averaged across all 1251 queries.

Method	Expected tokens	Savings
ripgrep + read file	45,692	baseline
semble	566	98% fewer

Recall at fixed token budgets

A relevant file is "covered" once any retrieved unit comes from it.

Method	500	1k	2k	4k	8k	16k	32k
semble	0.685	0.849	0.938	0.976	0.991	0.996	0.996
ripgrep + read file	0.001	0.008	0.037	0.088	0.212	0.379	0.583

Methodology

semble returns the top-50 ranked chunks. ripgrep+read splits the query into keywords (dropping stopwords and short words), runs rg --fixed-strings --ignore-case for each keyword, then reads matched files in full ranked by how many distinct keywords they contain. Both methods search the same set of file types and ignored directories. Tokens are counted with cl100k_base via tiktoken. A relevant file is "covered" once any retrieved unit overlaps its annotated span.

By language

NDCG@10 per language, sorted by CodeRankEmbed Hybrid (CRE in the table). Best score per row is bolded.

Language	semble	CRE Hybrid	CRE	ColGREP	grepai	probe	ripgrep
scala	0.909	0.922	0.845	0.765	0.330	0.392	0.180
cpp	0.915	0.913	0.846	0.626	0.731	0.375	0.126
ruby	0.909	0.909	0.769	0.708	0.643	0.382	0.230
elixir	0.894	0.905	0.869	0.808	0.669	0.412	0.134
javascript	0.917	0.903	0.920	0.823	0.675	0.588	0.176
zig	0.913	0.901	0.807	0.474	0.755	0.369	0.000
csharp	0.885	0.889	0.743	0.614	0.277	0.392	0.117
go	0.895	0.884	0.676	0.785	0.722	0.410	0.133
python	0.867	0.880	0.794	0.777	0.634	0.488	0.202
php	0.858	0.874	0.758	0.663	0.402	0.340	0.123
swift	0.860	0.873	0.721	0.710	0.429	0.280	0.160
bash	0.825	0.852	0.892	0.706	0.723	0.226	0.000
lua	0.823	0.847	0.803	0.798	0.699	0.336	0.000
java	0.849	0.841	0.706	0.641	0.386	0.536	0.198
kotlin	0.821	0.830	0.670	0.637	0.478	0.335	0.166
rust	0.856	0.827	0.627	0.662	0.519	0.242	0.162
c	0.741	0.806	0.706	0.676	0.555	0.384	0.000
haskell	0.765	0.771	0.776	0.683	0.483	0.313	0.000
typescript	0.706	0.708	0.545	0.430	0.394	0.354	0.128
overall	0.854	0.862	0.765	0.693	0.561	0.387	0.126

Ablations

raw returns retrieval scores directly; + ranking feeds them through semble's hybrid ranker.

Retrieval	Raw	+ ranking
BM25	0.675	0.834
potion-code-16M	0.650	0.821
BM25 + potion-code-16M	—	0.854

By query category

Mode	Architecture	Semantic	Symbol
BM25 raw	0.628	0.676	0.719
potion-code-16M raw	0.626	0.666	0.629
semble BM25 (+ ranking)	0.770	0.819	0.957
semble potion-code-16M (+ ranking)	0.757	0.808	0.943
semble hybrid	0.802	0.846	0.958

Dataset

~1,250 queries over 63 repositories in 19 languages, grouped into three categories:

Category	Queries	What it tests
semantic	711	Code that implements a specific behavior or concept
architecture	343	Design decisions, module boundaries, structural patterns
symbol	204	Named entity lookup (function, class, type, variable)

Notes

Languages: three repos per language (nine for Python): bash, C, C++, C#, Elixir, Go, Haskell, Java, JavaScript, Kotlin, Lua, PHP, Python, Ruby, Rust, Scala, Swift, TypeScript, Zig. Repos are pinned by revision in repos.json.

How the benchmark was built: queries and ground-truth relevance labels are generated by Claude Sonnet 4.6. The same model is used as LLM-as-judge to verify label quality.

Methods

• ripgrep: fast regex search over files, included as a raw keyword-match baseline.
• probe: BM25 keyword ranking backed by tree-sitter parse trees. No persistent index; scans on the fly.
• ColGREP: late-interaction code retrieval built on next-plaid with the LateOn-Code-edge model.
• grepai: semantic search using nomic-embed-text (137M params) via a local Ollama daemon.
• CodeRankEmbed: 137M-param transformer embedding model for code retrieval. CodeRankEmbed Hybrid fuses its dense scores with BM25.
• semble: this library. potion-code-16M static embeddings + BM25 + the semble reranking stack.

Excluded methods

Two tools were considered but not included in the benchmark:

• codanna: symbol-level semantic search with fastembed. Excluded because it does not support Haskell, Bash, Zig, Scala, Elixir, or Ruby (6 of the 19 benchmark languages, covering 20 of 63 repos (~38% of tasks)).
• claude-context: retrieval-augmented code search using OpenAI embeddings and a vector database. Excluded because it requires a paid OpenAI API key and a running vector-DB service.

Running the benchmarks

Repos are pinned in repos.json and cloned into ~/.cache/semble-bench:

uv run python -m benchmarks.sync_repos          # clone / update
uv run python -m benchmarks.sync_repos --check  # verify only

All tools run CPU-only. semble uses minishlab/potion-code-16M; CodeRankEmbed uses nomic-ai/CodeRankEmbed (137M params). The speed benchmark touches one repo per language with a cold-start index and 5 query runs per repo.

semble

uv run python -m benchmarks.run_benchmark
uv run python -m benchmarks.run_benchmark --repo fastapi --repo axios
uv run python -m benchmarks.run_benchmark --language python

Full runs write to benchmarks/results/semble-hybrid-<sha12>.json.

Speed benchmark

uv run python -m benchmarks.speed_benchmark

Writes to benchmarks/results/speed-<sha12>.json.

Ablations

uv run python -m benchmarks.baselines.ablations
uv run python -m benchmarks.baselines.ablations --mode bm25
uv run python -m benchmarks.baselines.ablations --mode semble-semantic

probe

Needs probe on $PATH (npm install -g @buger/probe).

uv run python -m benchmarks.baselines.probe
uv run python -m benchmarks.baselines.probe --repo fastapi --repo axios

grepai

Needs grepai on $PATH and Ollama running with nomic-embed-text pulled:

ollama pull nomic-embed-text

uv run python -m benchmarks.baselines.grepai
uv run python -m benchmarks.baselines.grepai --repo fastapi --repo axios

Large repos take several minutes to index. Use --timeout <seconds> (default 120) for repos with many files:

uv run python -m benchmarks.baselines.grepai --timeout 1800 --output results.json

The --output flag enables resume mode: already-completed repos are skipped on restart.

ripgrep

Needs rg on $PATH (brew install ripgrep / apt install ripgrep).

uv run python -m benchmarks.baselines.ripgrep
uv run python -m benchmarks.baselines.ripgrep --no-fixed-strings

ColGREP

Needs the colgrep binary on $PATH.

uv run python -m benchmarks.baselines.colgrep
uv run python -m benchmarks.baselines.colgrep --repo fastapi --repo axios

Runs with --code-only everywhere except bash repos (bash-it, bats-core, nvm), which use --no-code-only because ColGREP's code filter excludes .sh/.bash files.

CodeRankEmbed

Requires the benchmark extra (uv sync --extra benchmark).

uv run python -m benchmarks.baselines.coderankembed
uv run python -m benchmarks.baselines.coderankembed --mode semantic

Context-efficiency benchmark

Requires the benchmark extra (uv sync --extra benchmark) and rg on $PATH.

# Recall vs. token-budget across all queries; plots automatically.
uv run python -m benchmarks.token_efficiency recall
uv run python -m benchmarks.token_efficiency recall --repo fastapi

# Regenerate the plot from a saved recall payload.
uv run python -m benchmarks.token_efficiency plot

Writes benchmarks/results/token-efficiency-<sha12>.json and assets/images/token_efficiency.png.

Plots

uv run python -m benchmarks.plot

Writes speed_vs_ndcg_cold.png and speed_vs_ndcg_warm.png to assets/images/.