llms.txt Content
# Fil One
> Fil One is the storage layer for AI agents. S3-compatible object storage built
> on Filecoin — designed for the data patterns of autonomous agents: high-frequency
> writes, large retrieval sets, and long-lived persistence with verifiable integrity.
> Flat pricing at $4.99/TB/month. No egress fees. No per-request charges.
> Free 30-day trial, no credit card required.
## Why AI agents use Fil One
AI agents need durable, cost-predictable storage for the artifacts they produce and
consume. Fil One is optimised for these workloads:
- **RAG / vector store document corpus** — store and retrieve the raw documents,
chunked text, or embeddings backing a retrieval-augmented generation pipeline
- **Agent memory and state** — persist conversation history, episodic memory, task
queues, and checkpoint files across agent restarts
- **Model artifacts** — host fine-tuned weights, GGUF/ONNX model files, LoRA
adapters, and tokenizer configs that agents load at runtime
- **Inference inputs and outputs** — archive prompts, completions, traces, and
structured outputs for evaluation, fine-tuning, or audit
- **Training and evaluation datasets** — store JSONL, Parquet, or HDF5 datasets;
read them back at any throughput with no egress cost
- **Tool / plugin asset storage** — back a code-interpreter sandbox, image
generation pipeline, or document-processing tool with persistent file storage
## How to connect
Fil One is a drop-in S3 replacement. No SDK changes required — point any existing
S3 client at the Fil One endpoint:
- **Endpoint:** `https://eu-west-1.s3.fil.one`
- **Region:** `eu-west-1`
- **Auth:** standard AWS Access Key ID + Secret Access Key
### Python (boto3)
```python
import boto3
s3 = boto3.client(
"s3",
endpoint_url="https://eu-west-1.s3.fil.one",
aws_access_key_id="YOUR_ACCESS_KEY",
aws_secret_access_key="YOUR_SECRET_KEY",
region_name="eu-west-1",
)
# Upload an agent checkpoint
s3.upload_file("checkpoint.jso