BGE M3
Summary
BGE M3 is an open-weight language model from Baai, released on 30 Jan 2024. It accepts text and generates text, with a 8K-token context window and up to 1K output tokens per response. As of 17 Oct 2026, the lowest listed API price is $0.01 per million input tokens and Free per million output tokens (Novita AI), across 2 providers we track.
Multilingual embedding model combining dense, sparse, and multi-vector retrieval
Last verified 17 Oct 2026. Every figure on this page links to its source.
Specifications
| Developer | Baai |
|---|---|
| Release date | 30 Jan 2024 [models.dev] |
| Status | Generally available [models.dev] |
| Weights | Open weights [models.dev] |
| Context window | 8K (8,192 tokens) [models.dev] |
| Max output | 1K tokens [models.dev] |
| Input | Text [models.dev] |
| Output | Text [models.dev] |
| Reasoning mode | No [models.dev] |
| Tool calling | No [models.dev] |
BGE M3 API pricing by provider
USD per million tokens. Sorted by input price.
| Provider | Input | Output | Cached input | Context | Source |
|---|---|---|---|---|---|
| NVIDIA NIM baai/bge-m3 | Free | Free | – | 8K | models.dev, 17 Oct 2026 |
| Novita AI baai/bge-m3 | $0.01 | Free | – | 8K | models.dev, 17 Oct 2026 |
Frequently asked questions
What is the context window of BGE M3?
BGE M3 has a context window of 8,192 tokens (8K) and can generate up to 1,024 tokens in a single response, according to models.dev.
How much does the BGE M3 API cost?
BGE M3 is listed from $0.01 per million input tokens and Free per million output tokens (Novita AI). Prices last verified 17 Oct 2026.
Is BGE M3 open source?
BGE M3 is an open-weight model. Check the license terms before commercial use.
When was BGE M3 released?
Baai released BGE M3 on 30 Jan 2024, according to models.dev.
Which providers offer BGE M3?
We track BGE M3 on 2 providers: NVIDIA NIM, Novita AI.
Change history
- Added to the directory
- Now listed on Novita AI
- Now listed on NVIDIA NIM