Decision models · BAAI

BGE Reranker v2 M3

Task: Reranking search results
Returns: Probability or score
Precision tier: 72 / 100
Size: 570M
Max input: 8K tokens
License: apache-2.0
Released: 2024-03

What it decides

Returns: A relevance score for each query and document pair.
Checks: Query and document pairs
Languages: Multilingual
Precision: 72 / 100 · MTEB-R 57.0, MMTEB-R 58.4 (Qwen3-Reranker card); the most downloaded open reranker.
Clarity: Probability or score (15 of 15 clarity points)
How it runs: Scores query and document pairs · reads the input once and writes nothing, so it answers in milliseconds.
License: apache-2.0 (Open (Apache, MIT))

Source: model card on Hugging Face (opens in a new tab) · How decision models are rated

How to run it

Quantized files · GGUF

Works in llama.cpp (llama-server --rerank), Lemonade and other llama.cpp servers with a rerank endpoint. No sign-in needed.

huggingface.co/gpustack/bge-reranker-v2-m3-GGUF (opens in a new tab)

llama-server -hf gpustack/bge-reranker-v2-m3-GGUF --rerank

Original weights · Hugging Face

Full-precision files from BAAI, for Transformers, Sentence Transformers, Text Embeddings Inference, Infinity or vLLM.

huggingface.co/BAAI/bge-reranker-v2-m3 (opens in a new tab)

No sign-in needed.

Links checked 3 days ago against Hugging Face, Ollama and GitHub.

Apps that run BGE Reranker v2 M3

Our pick for each kind of computer, assuming the model fits in its memory. Pick your own GPU or Mac in the Model Finder for an exact answer.

Mac (Apple silicon)

Serves the GGUF build behind a /v1/rerank endpoint with llama-server --rerank.

Also good: Text Embeddings Inference, Infinity

Windows PC, NVIDIA GPU

Serves the GGUF build behind a /v1/rerank endpoint with llama-server --rerank.

Also good: Text Embeddings Inference, Sentence Transformers

Windows PC, AMD Radeon

Serves the GGUF build behind a /v1/rerank endpoint with llama-server --rerank.

Also good: Lemonade, Sentence Transformers

Linux, NVIDIA GPU

Serves the GGUF build behind a /v1/rerank endpoint with llama-server --rerank.

Also good: Text Embeddings Inference, Infinity

Linux server, 80 GB+ GPUs

Serves the GGUF build behind a /v1/rerank endpoint with llama-server --rerank.

Also good: Text Embeddings Inference, vLLM

App	Type	Opens it as	Systems	Engine	Start
llama.cpp	CLI + API	GGUF	macOSWindowsLinux	llama.cpp	`llama-server -hf gpustack/bge-reranker-v2-m3-GGUF:Q4_K_M --rerank`
vLLM	Server	Original weights	WSLWindowsLinux	vLLM	`vllm serve BAAI/bge-reranker-v2-m3`
SGLang	Server	Original weights	WSLWindowsLinux	SGLang	`python3 -m sglang.launch_server --model-path BAAI/bge-reranker-v2-m3`
Lemonade	Desktop app	GGUF	macOSWindowsLinux	llama.cpp, ONNX Runtime GenAI	`lemonade pull gpustack/bge-reranker-v2-m3-GGUF:Q4_K_M`
Text Embeddings Inference	Server	Original weights	macOSWSLWindowsLinux	TEI	`text-embeddings-router --model-id BAAI/bge-reranker-v2-m3 --port 8080`
Infinity	Server	Original weights	macOSWSLWindowsLinux	PyTorch, ONNX, CTranslate2	`infinity_emb v2 --model-id BAAI/bge-reranker-v2-m3`
LocalAI	Server	Original weights	macOSWSLWindowsLinux	llama.cpp, rerankers, Transformers	Get LocalAI
Sentence Transformers	Library	Original weights	macOSWindowsLinux	PyTorch	`python -c "from sentence_transformers import CrossEncoder; print(CrossEncoder('BAAI/bge-reranker-v2-m3').predict([('query', 'document')]))"`

Which apps can open a model depends on the files published for it: GGUF for llama.cpp-based apps, MLX for Apple’s engine, the original weights for vLLM and SGLang. Compare every app.

Memory and speed

Each precision on an NVIDIA GeForce RTX 4090, for one decision on a 512-token input. The best hardware table below covers other GPUs and Macs.

Format	Memory needed	Time per decision	Decisions per second	Quality	Download
`Q8_0`	1.1 GB	9.3 ms	107/s	72Great quality (A) · 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`F16`	1.7 GB	10 ms	99/s	72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab) Original weights (opens in a new tab)

One request at a time. Servers that batch (Text Embeddings Inference, vLLM, Infinity) handle many more decisions per second.

Best hardware for BGE Reranker v2 M3

The least expensive GPU in our catalog that runs it without offloading is the Intel Arc A380 (launch price $139) at F16.

Quality	Hardware	Quant	Memory	Time per decision	Runs
72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	AMD Radeon RX 7900 XTX	`F16`	1.9 GB / 24.0 GB	8.8 ms	Runs great
72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	AMD Radeon RX 9070 XT	`F16`	1.9 GB / 16.0 GB	10 ms	Runs great
72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	AMD Ryzen AI Max+ 395 (128 GB)	`F16`	1.9 GB / 96.0 GB	14 ms	Runs great
72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	Apple M1 Max (64 GB)	`F16`	1.9 GB / 48.0 GB	61 ms	Runs great
72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	Apple M2 Ultra (192 GB)	`F16`	1.9 GB / 144 GB	26 ms	Runs great
72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	Apple M3 Ultra (256 GB)	`F16`	1.9 GB / 192 GB	25 ms	Runs great
72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	Apple M3 Ultra (512 GB)	`F16`	1.9 GB / 384 GB	25 ms	Runs great
72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	Apple M4 Max (128 GB)	`F16`	1.9 GB / 96.0 GB	36 ms	Runs great
72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	Apple M4 Max (64 GB)	`F16`	1.9 GB / 48.0 GB	36 ms	Runs great
72Great quality (A) · 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking.	Apple M4 Pro (24 GB)	`F16`	1.9 GB / 16.1 GB	68 ms	Runs great

Check it on your own hardware

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.