Chain subnets into workflows.
One API call.

Compose multi-subnet pipelines in a single request. No centralized aggregator can replicate this.

τ

Collect Train Serve

Submit a topic, get back a trained model serving on your endpoint. Three subnets. One call.

Collect

SN13 Data Universe

Train

SN56 Gradients

Serve

SN64 Chutes
# one call: collect data → fine-tune → deploy
resp = requests.post(
    "https://api.opentau.ai/v1/pipelines",
    headers={"Authorization": "Bearer otau_your-key"},
    json={
        "pipeline": [
            {
                "step": "collect",
                "subnet": "sn13-data-universe",
                "params": {"topic": "solidity security", "count": 10000}
            },
            {
                "step": "train",
                "subnet": "sn56-gradients",
                "params": {"base_model": "llama-3.1-8b", "method": "qlora"}
            },
            {
                "step": "serve",
                "subnet": "sn64-chutes",
                "params": {"endpoint_name": "my-auditor"}
            }
        ]
    }
)
# → pipeline_id for async tracking
τ

Generate Verify

Every AI output automatically verified for authenticity. Trust built into the pipeline.

Generate

SN58/64 Inference

Verify

SN34 BitMind
# generate image + verify authenticity
resp = requests.post(
    "https://api.opentau.ai/v1/pipelines",
    headers={"Authorization": "Bearer otau_your-key"},
    json={
        "pipeline": [
            {
                "step": "generate",
                "subnet": "sn64-chutes",
                "params": {"prompt": "product photo, white bg"}
            },
            {
                "step": "verify",
                "subnet": "sn34-bitmind",
                "params": {"attach_certificate": True}
            }
        ]
    }
)
# → image + authenticity certificate
τ

Research Reason

Grounded, private, verifiable AI research. Search + data + synthesis, all routed through confidential compute.

Search

SN22 Desearch

Synthesize

SN13 + SN50

Reason

SN4 Targon (TEE)
# private research pipeline with confidential compute
resp = requests.post(
    "https://api.opentau.ai/v1/pipelines",
    headers={
        "Authorization": "Bearer otau_your-key",
        "X-Confidential": "true"
    },
    json={
        "pipeline": [
            {
                "step": "search",
                "subnet": "sn22-desearch",
                "params": {"query": "MEV protection mechanisms 2026"}
            },
            {
                "step": "synthesize",
                "subnet": "sn13-data-universe",
                "params": {"enrich": True}
            },
            {
                "step": "reason",
                "subnet": "sn4-targon",
                "params": {"model": "deepseek-r1", "confidential": True}
            }
        ]
    }
)
# → grounded analysis, private, verifiable

Build your own pipeline

Free tier includes pipeline access. Start chaining subnets today.

Get Your Free API Key →