All tasks / Models
F2LLM-v2-14B
Open-weights Apache-2.0 multilingual text embedding models from codefuse-ai, from 80M to 14B parameters.
- Ant Group (codefuse-ai)
- Open weights
- Released Mar 2026
Where it ranks
- AEmbeddings & retrievalModels · Local · 68.7% on MTEB(Multilingual, v2)
Facts
- Vendor
- Ant Group (codefuse-ai)
- Family
- F2LLM
- Released
- Mar 2026
- Weights
- Open: download and run it yourself
- License
- Apache-2.0
- Parameters
- 13.99B
- GPU memory to run locally
- ~28 GB (Derived: 13.99B parameters x 2 bytes = ~28 GB of bf16 weights; 8B, 4B, 1.7B, 0.6B and smaller sizes exist.)
- Access
- Free (open weights)
- Last verified
- 19 Oct 2026
Benchmark results
| Benchmark | Result | Entry on the leaderboard | As of | Source |
|---|---|---|---|---|
| MTEB(eng, v2) | 73.1% | codefuse-ai/F2LLM-v2-14BMean (Task) over all 41 tasks; board rank 11 (Borda); board zero-shot 43%; read from the leaderboard's public API. | 19 Oct 2026 | mteb-leaderboard-backend.hf.space |
| MTEB(Multilingual, v2) | 68.7% | codefuse-ai/F2LLM-v2-14BMean (Task) over all 131 tasks; board rank 8 (Borda); board zero-shot 88%; read from the leaderboard's public API. | 19 Oct 2026 | mteb-leaderboard-backend.hf.space |
Where you can use it
No tracked tool lists this model yet; as open weights it runs in local runtimes.