BAAI · Text
BGE models: Versions, specs, GPU requirements & hosting
BAAI BGE separates general embedding and reranking checkpoints for retrieval pipelines that need multilingual dense, sparse and multi-vector representations.
Overview and history
BGE is an official BAAI model family represented here through independently verified release, checkpoint, license, runtime and hardware evidence. BAAI BGE separates general embedding and reranking checkpoints for retrieval pipelines that need multilingual dense, sparse and multi-vector representations. Current public releases are grouped by the Embedding, Reranking tracks so sibling sizes and specialist branches are not presented as false chronological generations.
FusionPointAI currently tracks 6 certified public BGE releases. The catalogue preserves release dates and evaluates Current, Recent, Previous and Legacy status within each model track rather than by database insertion order.
Embedding and reranking tracks
Retrieval-focused checkpoints
Sentence Transformers ecosystem
Architecture and model range
This Embedding and reranking family has 6 source-certified public releases. Architecture, parameter counts, and pipeline details are shown only when the official source publishes them; language-model-only context and KV-cache assumptions are not applied.
Public releases6
Model tracks2
Current generationBGE Reranker v2 M3
Last verified2026-09-03T00:00:00.000000Z
Generations, tracks and variants
Embedding
cci3-hq-classifier
Recent generationindustrycorpus2-datarater
Recent generationindustrycorpus2-classifier
Recent generationBGE-M3
Current generationReranking
bge-reranker-v2-m3
Current generation 2 sibling variantsDeployment ecosystem
Choose a specific BGE release and certified checkpoint before sizing hardware. FusionPointAI rejects unknown runtime configurations and applies workload-specific memory rules before separating compatible hardware from rentable inventory.
Validate the exact checkpoint, license, runtime, precision, input and output dimensions, batch size, and GPU memory. Resource sizing uses rules appropriate to Embedding and reranking workloads.
Common use cases
Private Embedding and reranking workloads
Controlled model evaluation
Production deployment planning
License overview
Licenses are certified per release, not inferred from the provider. Review the release page before commercial deployment because terms can differ between generations and variants.
Commercial-use status is a source-backed catalogue classification, not legal advice. Follow the linked official license and attribution requirements.
Release timeline
baai/bge/cci3-hq-classifier @ 69fd74e73f6550209a0912484b5321745e530be7ef134537a4c590d0b43558f2
baai/bge/industrycorpus2-classifier @ c08c6369b1a46f2795b3a7eaf32740abe37d6665d83688e0b67796b0ed789d7f
baai/bge/industrycorpus2-datarater @ fe3efe946494d55dac01d533608053ccbdf58a93fe7b56e9c57058172df0a1de
baai/bge/bge-bge-reranker-v2-m3 @ 1b356f57ae4470f30f8fa3f95c8f0a253e4032038a1dd5ff51d9c463d43f0e55
baai/bge/bge-reranker-v2-m3 @ a0f5b78dcaca3b947f9c3996ca54b10da5328cdf462272540df0913655985770
BAAI/bge-m3 @ 2024-09-14T00:26:03Z
BAAI/bge-reranker-v2-m3 @ 2024-09-13T06:37:24Z
Frequently asked questions
Which BGE release is current?
Current releases are BGE-M3, BGE Reranker v2 M3, bge-reranker-v2-m3, evaluated within their respective tracks.
Can BGE be self-hosted?
Only releases with a verified official checkpoint and supported runtime configuration are presented as deployable.
Which GPU should run BGE?
GPU selection depends on the exact variant, checkpoint precision, context length and concurrency; use the release hardware recommendations rather than the family name alone.
Official references
Last verified: 2026-09-03