10 actions in 9 pieces of software · ranked by words, meaning and measured trust · 1085 ms

1. 1SciJudge-4Bverified and signedpassport

   modelrecord40 downloads/month

   • text-generation

     Task: text-generation

2. 2Qwen3-Reranker-4Bverified and signedpassport

   modelrecord2,412,394 downloads/month

   • text-ranking

     Task: text-ranking

3. 3t5gemma-2-4b-4bverified and signedpassport

   modelrecord14,503 downloads/month

   • image-text-to-text

     Task: image-text-to-text

4. 4decider-4bverified and signedpassport

   modelrecord29,629 downloads/month

   • text-classification

     Task: text-classification

5. 5ibm-granite/granite-vision-4.1-4bverified and signedpassport

   modelrecord

   • ibm-granite/granite-vision-4.1-4b

     Granite Vision 4.1 4B is a vision-language model (VLM) that delivers frontier-level performance on structured document extraction tasks — chart extraction, table extraction, and semantic key-value pair extraction — in a compact 4B parameter footprint

6. 6Nemotron-Content-Safety-Reasoning-4B (Experimental)verified and signedpassport

   modelrecord

   • natural-language-understanding

     Task: Natural Language Understanding

   • natural-language-processing

     Task: Natural Language Processing

7. 7black-forest-labs/flux-2-klein-4b-base-loraverified and signedpassport

   modelrecord

   • black-forest-labs/flux-2-klein-4b-base-lora

     A version of FLUX.2 [klein] 4B-base that supports fast fine-tuned lora inference

8. 8NavGPT3-4Bverified and signedpassport

   modelrecord154 downloads/month

   • robotics

     Task: robotics

9. 9Tell-4Bdemo or tutorialno passport yet

   MCP serverrecord

   • tell_4b_demo_decide

     Answer one Tell question: state + question + options -> one label. Returns: A markdown summary, the per-label probabilities, and the full result, object (ranking, token count, exact prompt) for inspection.