Why did we open-source our inference engine? Read the post

Build manufacturing agents with open source models

Superlinked gives your agent one API to read manuals, match sensor events to maintenance records, extract equipment details and return cited findings.

Get started
See how our example agent reconstructed the NTSB detector sequence

Your reliability agent

Join sensor readings, images and technical findings.

Investigation question What did the last three detectors record, and who got the alerts?
Evidence summary 38°F, 103°F to the Help Desk, 253°F to the cab.

Your operations context

Sensor history, inspection records, images and equipment guidance.

PDF NTSB digest Detector sequence
SENSOR Hot-bearing detector readings 38°F · 103°F · 253°F
IMAGE Wayside cameras Salem fire
RULE Alert routing Desk · crew
Browse all tasks
from sie_sdk import SIEClient
client = SIEClient(
api_key="sk-sie-…",
base_url="https://api.superlinked.com",
)
# Parse a document → clean markdown, tables and layout kept.
result = client.extract(
"modeldocling",
{"document": filePDFntsb-east-p-ohio-illustrated-digest.pdfbrowse},
)
print(result["data"]["markdown"])
document_bytes=$(base64 < 'ntsb-east-p-ohio-illustrated-digest.pdf' | tr -d '\n')
curl https://api.superlinked.com/v1/extract/docling \
  -H "Authorization: Bearer sk-sie-…" \
  -H "Content-Type: application/json" \
  -d "{\"items\":[{\"document\":{\"data\":\"$document_bytes\",\"format\":\"pdf\"}}]}"
Build the "Doc to Markdown" capability into my app using the Superlinked Inference Engine (SIE).

Context
- SIE is an OpenAI-style inference API. Python SDK: `from sie_sdk import SIEClient`; TypeScript: `@superlinked/sie-sdk`.
- Base URL: https://api.superlinked.com (or my regional endpoint). Auth: Bearer key from env `SIE_API_KEY` (never hard-code it).
- Model: docling (SIE primitive: /extract). Keep the model id configurable.

Task
- Input: an uploaded document (PDF / Office / scan).
- Behaviour: return clean markdown for the document, preserving tables and reading order
- Call the selected SIE primitive once per request and map the response into your domain type.

Deliverables
- A typed client wrapper, an application-level function for this task, error handling for timeouts/empty input, and unit tests with a stubbed client.
- Wire it into my existing stack (ask me which framework if unclear) and add a short usage example.
Input PDF spreadntsb-east-p-ohio-illustrated-digest.pdf
NTSB illustrated digest spread with detector readings, map and surveillance images
from sie_sdk import SIEClient
client = SIEClient(
api_key="sk-sie-…",
base_url="https://api.superlinked.com",
)
result = client.extract(
"modelfastino/gliner2-large-v1",
{"text": "textAt 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted."},
labels=[
"entity typelocation",
"entity typeevent time",
"entity typebearing",
"entity typedegrees above ambient",
"entity typealert status",
"entity typealert recipient",
"entity typecamera observation",
],
)
for span in result["entities"]:
print(span["label"], span["text"])
import { SIEClient } from '@superlinked/sie-sdk';

const client = new SIEClient('https://api.superlinked.com', {
  apiKey: 'sk-sie-…',
});

const result = await client.extract(
  'fastino/gliner2-large-v1',
  { text: "At 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted." },
  { labels: ["location","event time","bearing","degrees above ambient","alert status","alert recipient","camera observation"] },
);
console.log(result.entities);
curl https://api.superlinked.com/v1/extract/fastino%2Fgliner2-large-v1 \
  -H "Authorization: Bearer sk-sie-…" \
  -H "Content-Type: application/json" \
  -d "{\"items\":[{\"text\":\"At 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted.\"}],\"params\":{\"labels\":[\"location\",\"event time\",\"bearing\",\"degrees above ambient\",\"alert status\",\"alert recipient\",\"camera observation\"]}}"
Build the "Named entities" capability into my app using the Superlinked Inference Engine (SIE).

Context
- SIE is an OpenAI-style inference API. Python SDK: `from sie_sdk import SIEClient`; TypeScript: `@superlinked/sie-sdk`.
- Base URL: https://api.superlinked.com (or my regional endpoint). Auth: Bearer key from env `SIE_API_KEY` (never hard-code it).
- Model: fastino/gliner2-large-v1 (SIE primitive: /extract). Keep the model id configurable.

Task
- Input: a block of text.
- Behaviour: return the named entities found in the text with their types
- Call the selected SIE primitive once per request and map the response into your domain type.

Deliverables
- A typed client wrapper, an application-level function for this task, error handling for timeouts/empty input, and unit tests with a stubbed client.
- Wire it into my existing stack (ask me which framework if unclear) and add a short usage example.
import numpy as np
from sie_sdk import SIEClient
client = SIEClient(
api_key="sk-sie-…",
base_url="https://api.superlinked.com",
)
query = "queryHow did the L1 bearing temperature and alert status change across Sebring, Salem, and East P., Ohio, and what did the NTSB report happened next?"
documents = [
"document 1## When the bearing failed",
"document 2The East P., Ohio derailment began when an overheated bearing burned off the accident hopper car. Here is what you need to know about bearings, hot bearing detectors (HBDs), and what happened when the accident hopper car passed over the final three HBDs of the journey in the Ohio towns of Sebring, Salem, and East P., Ohio.",
"document 3A bearing (A) fits around the end of a railcar axle. The part of the axle extending into the bearing is called the journal (B) . The bearing's roller components allow the journal to turn within the bearing with little friction. However, when a bearing fails, the resulting friction can be catastrophic. In East P., Ohio, the journal's cylindrical shape was worn down to a cone (C) . ( See box below .)",
"document 4Overheated wheel bearings are the most common mechanical or electrical cause of rail accidents. Because overheated bearings are so dangerous, railroad tracks are equipped with HBDs, devices located along the track that measure the bearing's temperature relative to the ambient temperature (D) . Ambient temperature means the temperature outside. For example, the ambient temperature at Sebring, Ohio, was measured by a thermometer at the nearby signal bungalow.",
"document 5Bearings can go from below-alarm temperature to seized in less than a minute. In this accident, however, fire was first photographed near the L1 bearing by a surveillance camera in Salem, Ohio (right), 39 minutes before the train derailed in East P., Ohio.",
"document 6The bearing that caused the accident hopper car to derail was its L1 bearing, which is on the right side of the car, facing the direction of travel.",
"document 7Railcar axles are numbered starting from the end of the railcar with the brake wheel (the 'B end'). When the B end is in the front of the car (when it faces the direction of travel) the 'left-side' wheels and bearings are on the right.",
"document 8## The undetected hot bearing burns off",
"document 91. HBDs (D) measure the temperature on the outside of the bearing cup. When a bearing heats up inside, it may take 30-60 minutes for the HBD to sense it from the outside. The HBD in Salem measured a temperature of only 103° above ambient, despite the fact that the bearing was on fire. So, it generated an alert to the Wayside Help Desk, not an alarm to the train crew.",
"document 102. By the time the alarm sounded at the East P., Ohio HBD, the journal was worn down to a cone, and it was too late. The hopper car and 37 others derailed as the train's emergency braking system activated.",
"document 11## On fire but not 'critical'",
"document 12At 7:37 p.m. , at the Sebring HBD, the reading for the L1 bearing was only 38°F above ambient -not high enough to trigger an alert. The R1 bearing measured 20°F above ambient. But the L1 bearing was hotter.",
"document 13At 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted.",
"document 14About 8:52 p.m ., the train went over the East P., Ohio HBD. Less than a minute later, the HBD recorded a temperature of 253°F above ambient at the L1 bearing; bearing R1 remained at 20°F above ambient. The HBD immediately transmitted a critical alarm, which was broadcast in the locomotive cab.",
"document 15Under NS rules, if a train's crew hears a critical alarm, they must stop the train right away and inspect the wheelset that triggered the alarm. The engineer began to slow the train before 8:54 p.m.",
"document 16But it was too late; the hopper car and 37 others derailed as the train's emergency braking system activated.",
"document 17The Wayside Help Desk in Atlanta, Georgia, monitors alerts from NS's HBDs. Critical alarms broadcast straight to the train crew. For noncritical alerts, the analyst monitors the train over HBDs it crosses later, paying attention to temperature trends. The single analyst on duty did not notice the Salem noncritical alert, as he dealt with higher-priority alerts. But if he did notice the alert, that would not have prevented the derailment. The analyst said that if he had seen the Salem HBD alert in real time, he would have followed NS rules and watched for alerts from HBDs it went over later. The next HBD the accident train went over was in East P., Ohio.",
"document 18In the East P., Ohio derailment, the alarm was received by the lead locomotive radio and broadcast over the in-cab speakers: ' Critical alarm, critical alarm, critical alarm. Norfolk Southern milepost 49.8, track 1 hotbox, axle 101, south rail. '",
]
items = [{"text": query}, *({"text": d} for d in documents)]
vecs = client.encode("modelQwen/Qwen3-Embedding-4B", items)
mat = np.array([v["dense"] for v in vecs])
mat = mat / np.linalg.norm(mat, axis=1, keepdims=True)
scores = mat[1:] @ mat[0] # cosine similarity to the query
for i in np.argsort(scores)[::-1]:
print(f"{scores[i]:.3f} {documents[i]}")
import { SIEClient } from '@superlinked/sie-sdk';

const client = new SIEClient('https://api.superlinked.com', {
  apiKey: 'sk-sie-…',
});

const query = "How did the L1 bearing temperature and alert status change across Sebring, Salem, and East P., Ohio, and what did the NTSB report happened next?";
const documents = [
  "## When the bearing failed",
  "The East P., Ohio derailment began when an overheated bearing burned off the accident hopper car. Here is what you need to know about bearings, hot bearing detectors (HBDs), and what happened when the accident hopper car passed over the final three HBDs of the journey in the Ohio towns of Sebring, Salem, and East P., Ohio.",
  "A bearing (A) fits around the end of a railcar axle. The part of the axle extending into the bearing is called the journal (B) . The bearing's roller components allow the journal to turn within the bearing with little friction. However, when a bearing fails, the resulting friction can be catastrophic. In East P., Ohio, the journal's cylindrical shape was worn down to a cone (C) . ( See box below .)",
  "Overheated wheel bearings are the most common mechanical or electrical cause of rail accidents. Because overheated bearings are so dangerous, railroad tracks are equipped with HBDs, devices located along the track that measure the bearing's temperature relative to the ambient temperature (D) . Ambient temperature means the temperature outside. For example, the ambient temperature at Sebring, Ohio, was measured by a thermometer at the nearby signal bungalow.",
  "Bearings can go from below-alarm temperature to seized in less than a minute. In this accident, however, fire was first photographed near the L1 bearing by a surveillance camera in Salem, Ohio (right), 39 minutes before the train derailed in East P., Ohio.",
  "The bearing that caused the accident hopper car to derail was its L1 bearing, which is on the right side of the car, facing the direction of travel.",
  "Railcar axles are numbered starting from the end of the railcar with the brake wheel (the 'B end'). When the B end is in the front of the car (when it faces the direction of travel) the 'left-side' wheels and bearings are on the right.",
  "## The undetected hot bearing burns off",
  "1. HBDs (D) measure the temperature on the outside of the bearing cup. When a bearing heats up inside, it may take 30-60 minutes for the HBD to sense it from the outside. The HBD in Salem measured a temperature of only 103° above ambient, despite the fact that the bearing was on fire. So, it generated an alert to the Wayside Help Desk, not an alarm to the train crew.",
  "2. By the time the alarm sounded at the East P., Ohio HBD, the journal was worn down to a cone, and it was too late. The hopper car and 37 others derailed as the train's emergency braking system activated.",
  "## On fire but not 'critical'",
  "At 7:37 p.m. , at the Sebring HBD, the reading for the L1 bearing was only 38°F above ambient -not high enough to trigger an alert. The R1 bearing measured 20°F above ambient. But the L1 bearing was hotter.",
  "At 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted.",
  "About 8:52 p.m ., the train went over the East P., Ohio HBD. Less than a minute later, the HBD recorded a temperature of 253°F above ambient at the L1 bearing; bearing R1 remained at 20°F above ambient. The HBD immediately transmitted a critical alarm, which was broadcast in the locomotive cab.",
  "Under NS rules, if a train's crew hears a critical alarm, they must stop the train right away and inspect the wheelset that triggered the alarm. The engineer began to slow the train before 8:54 p.m.",
  "But it was too late; the hopper car and 37 others derailed as the train's emergency braking system activated.",
  "The Wayside Help Desk in Atlanta, Georgia, monitors alerts from NS's HBDs. Critical alarms broadcast straight to the train crew. For noncritical alerts, the analyst monitors the train over HBDs it crosses later, paying attention to temperature trends. The single analyst on duty did not notice the Salem noncritical alert, as he dealt with higher-priority alerts. But if he did notice the alert, that would not have prevented the derailment. The analyst said that if he had seen the Salem HBD alert in real time, he would have followed NS rules and watched for alerts from HBDs it went over later. The next HBD the accident train went over was in East P., Ohio.",
  "In the East P., Ohio derailment, the alarm was received by the lead locomotive radio and broadcast over the in-cab speakers: ' Critical alarm, critical alarm, critical alarm. Norfolk Southern milepost 49.8, track 1 hotbox, axle 101, south rail. '",
];

const items = [{ text: query }, ...documents.map((text) => ({ text }))];
const vecs = await client.encode('Qwen/Qwen3-Embedding-4B', items);
const [q, ...d] = vecs.map((v) => v.dense);
const cosine = (a: Float32Array, b: Float32Array) => {
  const dot = a.reduce((s, x, i) => s + x * b[i], 0);
  return dot / (Math.hypot(...a) * Math.hypot(...b));
};
const ranked = documents
  .map((doc, i) => ({ doc, score: cosine(q!, d[i]!) }))
  .sort((a, b) => b.score - a.score);
console.log(ranked);
# Encode the query + candidates in one batch, then rank by cosine.
curl https://api.superlinked.com/v1/encode/Qwen%2FQwen3-Embedding-4B \
  -H "Authorization: Bearer sk-sie-…" \
  -H "Content-Type: application/json" \
  -d "{\"items\":[{\"text\":\"How did the L1 bearing temperature and alert status change across Sebring, Salem, and East P., Ohio, and what did the NTSB report happened next?\"},{\"text\":\"## When the bearing failed\"},{\"text\":\"The East P., Ohio derailment began when an overheated bearing burned off the accident hopper car. Here is what you need to know about bearings, hot bearing detectors (HBDs), and what happened when the accident hopper car passed over the final three HBDs of the journey in the Ohio towns of Sebring, Salem, and East P., Ohio.\"},{\"text\":\"A bearing (A) fits around the end of a railcar axle. The part of the axle extending into the bearing is called the journal (B) . The bearing's roller components allow the journal to turn within the bearing with little friction. However, when a bearing fails, the resulting friction can be catastrophic. In East P., Ohio, the journal's cylindrical shape was worn down to a cone (C) . ( See box below .)\"},{\"text\":\"Overheated wheel bearings are the most common mechanical or electrical cause of rail accidents. Because overheated bearings are so dangerous, railroad tracks are equipped with HBDs, devices located along the track that measure the bearing's temperature relative to the ambient temperature (D) . Ambient temperature means the temperature outside. For example, the ambient temperature at Sebring, Ohio, was measured by a thermometer at the nearby signal bungalow.\"},{\"text\":\"Bearings can go from below-alarm temperature to seized in less than a minute. In this accident, however, fire was first photographed near the L1 bearing by a surveillance camera in Salem, Ohio (right), 39 minutes before the train derailed in East P., Ohio.\"},{\"text\":\"The bearing that caused the accident hopper car to derail was its L1 bearing, which is on the right side of the car, facing the direction of travel.\"},{\"text\":\"Railcar axles are numbered starting from the end of the railcar with the brake wheel (the 'B end'). When the B end is in the front of the car (when it faces the direction of travel) the 'left-side' wheels and bearings are on the right.\"},{\"text\":\"## The undetected hot bearing burns off\"},{\"text\":\"1. HBDs (D) measure the temperature on the outside of the bearing cup. When a bearing heats up inside, it may take 30-60 minutes for the HBD to sense it from the outside. The HBD in Salem measured a temperature of only 103° above ambient, despite the fact that the bearing was on fire. So, it generated an alert to the Wayside Help Desk, not an alarm to the train crew.\"},{\"text\":\"2. By the time the alarm sounded at the East P., Ohio HBD, the journal was worn down to a cone, and it was too late. The hopper car and 37 others derailed as the train's emergency braking system activated.\"},{\"text\":\"## On fire but not 'critical'\"},{\"text\":\"At 7:37 p.m. , at the Sebring HBD, the reading for the L1 bearing was only 38°F above ambient -not high enough to trigger an alert. The R1 bearing measured 20°F above ambient. But the L1 bearing was hotter.\"},{\"text\":\"At 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted.\"},{\"text\":\"About 8:52 p.m ., the train went over the East P., Ohio HBD. Less than a minute later, the HBD recorded a temperature of 253°F above ambient at the L1 bearing; bearing R1 remained at 20°F above ambient. The HBD immediately transmitted a critical alarm, which was broadcast in the locomotive cab.\"},{\"text\":\"Under NS rules, if a train's crew hears a critical alarm, they must stop the train right away and inspect the wheelset that triggered the alarm. The engineer began to slow the train before 8:54 p.m.\"},{\"text\":\"But it was too late; the hopper car and 37 others derailed as the train's emergency braking system activated.\"},{\"text\":\"The Wayside Help Desk in Atlanta, Georgia, monitors alerts from NS's HBDs. Critical alarms broadcast straight to the train crew. For noncritical alerts, the analyst monitors the train over HBDs it crosses later, paying attention to temperature trends. The single analyst on duty did not notice the Salem noncritical alert, as he dealt with higher-priority alerts. But if he did notice the alert, that would not have prevented the derailment. The analyst said that if he had seen the Salem HBD alert in real time, he would have followed NS rules and watched for alerts from HBDs it went over later. The next HBD the accident train went over was in East P., Ohio.\"},{\"text\":\"In the East P., Ohio derailment, the alarm was received by the lead locomotive radio and broadcast over the in-cab speakers: ' Critical alarm, critical alarm, critical alarm. Norfolk Southern milepost 49.8, track 1 hotbox, axle 101, south rail. '\"}],\"params\":{\"output_types\":[\"dense\"]}}"
Build the "Search" capability into my app using the Superlinked Inference Engine (SIE).

Context
- SIE is an OpenAI-style inference API. Python SDK: `from sie_sdk import SIEClient`; TypeScript: `@superlinked/sie-sdk`.
- Base URL: https://api.superlinked.com (or my regional endpoint). Auth: Bearer key from env `SIE_API_KEY` (never hard-code it).
- Model: Qwen/Qwen3-Embedding-4B (SIE primitive: /encode). Keep the model id configurable.

Task
- Input: a query string plus a list of candidate documents.
- Behaviour: return the candidates ranked by semantic similarity to the query, with a score per candidate
- Encode the query and candidates in one batched /encode call, then rank by cosine similarity in the client. Do not call the API per candidate.

Deliverables
- A typed client wrapper, an application-level function for this task, error handling for timeouts/empty input, and unit tests with a stubbed client.
- Wire it into my existing stack (ask me which framework if unclear) and add a short usage example.
from sie_sdk import SIEClient
client = SIEClient(
api_key="sk-sie-…",
base_url="https://api.superlinked.com",
)
query = "queryHow did the L1 bearing temperature and alert status change across Sebring, Salem, and East P., Ohio, and what did the NTSB report happened next?"
documents = [
"document 1## When the bearing failed",
"document 2The East P., Ohio derailment began when an overheated bearing burned off the accident hopper car. Here is what you need to know about bearings, hot bearing detectors (HBDs), and what happened when the accident hopper car passed over the final three HBDs of the journey in the Ohio towns of Sebring, Salem, and East P., Ohio.",
"document 3A bearing (A) fits around the end of a railcar axle. The part of the axle extending into the bearing is called the journal (B) . The bearing's roller components allow the journal to turn within the bearing with little friction. However, when a bearing fails, the resulting friction can be catastrophic. In East P., Ohio, the journal's cylindrical shape was worn down to a cone (C) . ( See box below .)",
"document 4Overheated wheel bearings are the most common mechanical or electrical cause of rail accidents. Because overheated bearings are so dangerous, railroad tracks are equipped with HBDs, devices located along the track that measure the bearing's temperature relative to the ambient temperature (D) . Ambient temperature means the temperature outside. For example, the ambient temperature at Sebring, Ohio, was measured by a thermometer at the nearby signal bungalow.",
"document 5Bearings can go from below-alarm temperature to seized in less than a minute. In this accident, however, fire was first photographed near the L1 bearing by a surveillance camera in Salem, Ohio (right), 39 minutes before the train derailed in East P., Ohio.",
"document 6The bearing that caused the accident hopper car to derail was its L1 bearing, which is on the right side of the car, facing the direction of travel.",
"document 7Railcar axles are numbered starting from the end of the railcar with the brake wheel (the 'B end'). When the B end is in the front of the car (when it faces the direction of travel) the 'left-side' wheels and bearings are on the right.",
"document 8## The undetected hot bearing burns off",
"document 91. HBDs (D) measure the temperature on the outside of the bearing cup. When a bearing heats up inside, it may take 30-60 minutes for the HBD to sense it from the outside. The HBD in Salem measured a temperature of only 103° above ambient, despite the fact that the bearing was on fire. So, it generated an alert to the Wayside Help Desk, not an alarm to the train crew.",
"document 102. By the time the alarm sounded at the East P., Ohio HBD, the journal was worn down to a cone, and it was too late. The hopper car and 37 others derailed as the train's emergency braking system activated.",
"document 11## On fire but not 'critical'",
"document 12At 7:37 p.m. , at the Sebring HBD, the reading for the L1 bearing was only 38°F above ambient -not high enough to trigger an alert. The R1 bearing measured 20°F above ambient. But the L1 bearing was hotter.",
"document 13At 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted.",
"document 14About 8:52 p.m ., the train went over the East P., Ohio HBD. Less than a minute later, the HBD recorded a temperature of 253°F above ambient at the L1 bearing; bearing R1 remained at 20°F above ambient. The HBD immediately transmitted a critical alarm, which was broadcast in the locomotive cab.",
"document 15Under NS rules, if a train's crew hears a critical alarm, they must stop the train right away and inspect the wheelset that triggered the alarm. The engineer began to slow the train before 8:54 p.m.",
"document 16But it was too late; the hopper car and 37 others derailed as the train's emergency braking system activated.",
"document 17The Wayside Help Desk in Atlanta, Georgia, monitors alerts from NS's HBDs. Critical alarms broadcast straight to the train crew. For noncritical alerts, the analyst monitors the train over HBDs it crosses later, paying attention to temperature trends. The single analyst on duty did not notice the Salem noncritical alert, as he dealt with higher-priority alerts. But if he did notice the alert, that would not have prevented the derailment. The analyst said that if he had seen the Salem HBD alert in real time, he would have followed NS rules and watched for alerts from HBDs it went over later. The next HBD the accident train went over was in East P., Ohio.",
"document 18In the East P., Ohio derailment, the alarm was received by the lead locomotive radio and broadcast over the in-cab speakers: ' Critical alarm, critical alarm, critical alarm. Norfolk Southern milepost 49.8, track 1 hotbox, axle 101, south rail. '",
]
items = [{"id": str(i), "text": d} for i, d in enumerate(documents)]
ranked = client.score("modelQwen/Qwen3-Reranker-4B", {"text": query}, items)
for r in ranked["scores"]:
print(r["score"], documents[int(r["item_id"])])
import { SIEClient } from '@superlinked/sie-sdk';

const client = new SIEClient('https://api.superlinked.com', {
  apiKey: 'sk-sie-…',
});

const query = "How did the L1 bearing temperature and alert status change across Sebring, Salem, and East P., Ohio, and what did the NTSB report happened next?";
const documents = [
  "## When the bearing failed",
  "The East P., Ohio derailment began when an overheated bearing burned off the accident hopper car. Here is what you need to know about bearings, hot bearing detectors (HBDs), and what happened when the accident hopper car passed over the final three HBDs of the journey in the Ohio towns of Sebring, Salem, and East P., Ohio.",
  "A bearing (A) fits around the end of a railcar axle. The part of the axle extending into the bearing is called the journal (B) . The bearing's roller components allow the journal to turn within the bearing with little friction. However, when a bearing fails, the resulting friction can be catastrophic. In East P., Ohio, the journal's cylindrical shape was worn down to a cone (C) . ( See box below .)",
  "Overheated wheel bearings are the most common mechanical or electrical cause of rail accidents. Because overheated bearings are so dangerous, railroad tracks are equipped with HBDs, devices located along the track that measure the bearing's temperature relative to the ambient temperature (D) . Ambient temperature means the temperature outside. For example, the ambient temperature at Sebring, Ohio, was measured by a thermometer at the nearby signal bungalow.",
  "Bearings can go from below-alarm temperature to seized in less than a minute. In this accident, however, fire was first photographed near the L1 bearing by a surveillance camera in Salem, Ohio (right), 39 minutes before the train derailed in East P., Ohio.",
  "The bearing that caused the accident hopper car to derail was its L1 bearing, which is on the right side of the car, facing the direction of travel.",
  "Railcar axles are numbered starting from the end of the railcar with the brake wheel (the 'B end'). When the B end is in the front of the car (when it faces the direction of travel) the 'left-side' wheels and bearings are on the right.",
  "## The undetected hot bearing burns off",
  "1. HBDs (D) measure the temperature on the outside of the bearing cup. When a bearing heats up inside, it may take 30-60 minutes for the HBD to sense it from the outside. The HBD in Salem measured a temperature of only 103° above ambient, despite the fact that the bearing was on fire. So, it generated an alert to the Wayside Help Desk, not an alarm to the train crew.",
  "2. By the time the alarm sounded at the East P., Ohio HBD, the journal was worn down to a cone, and it was too late. The hopper car and 37 others derailed as the train's emergency braking system activated.",
  "## On fire but not 'critical'",
  "At 7:37 p.m. , at the Sebring HBD, the reading for the L1 bearing was only 38°F above ambient -not high enough to trigger an alert. The R1 bearing measured 20°F above ambient. But the L1 bearing was hotter.",
  "At 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted.",
  "About 8:52 p.m ., the train went over the East P., Ohio HBD. Less than a minute later, the HBD recorded a temperature of 253°F above ambient at the L1 bearing; bearing R1 remained at 20°F above ambient. The HBD immediately transmitted a critical alarm, which was broadcast in the locomotive cab.",
  "Under NS rules, if a train's crew hears a critical alarm, they must stop the train right away and inspect the wheelset that triggered the alarm. The engineer began to slow the train before 8:54 p.m.",
  "But it was too late; the hopper car and 37 others derailed as the train's emergency braking system activated.",
  "The Wayside Help Desk in Atlanta, Georgia, monitors alerts from NS's HBDs. Critical alarms broadcast straight to the train crew. For noncritical alerts, the analyst monitors the train over HBDs it crosses later, paying attention to temperature trends. The single analyst on duty did not notice the Salem noncritical alert, as he dealt with higher-priority alerts. But if he did notice the alert, that would not have prevented the derailment. The analyst said that if he had seen the Salem HBD alert in real time, he would have followed NS rules and watched for alerts from HBDs it went over later. The next HBD the accident train went over was in East P., Ohio.",
  "In the East P., Ohio derailment, the alarm was received by the lead locomotive radio and broadcast over the in-cab speakers: ' Critical alarm, critical alarm, critical alarm. Norfolk Southern milepost 49.8, track 1 hotbox, axle 101, south rail. '",
];
const items = documents.map((text, i) => ({ id: String(i), text }));
const ranked = await client.score('Qwen/Qwen3-Reranker-4B', { text: query }, items);
console.log(ranked.scores.map(({ itemId, score }) => ({ document: documents[Number(itemId)], score })));
curl https://api.superlinked.com/v1/score/Qwen%2FQwen3-Reranker-4B \
  -H "Authorization: Bearer sk-sie-…" \
  -H "Content-Type: application/json" \
  -d "{\"query\":{\"text\":\"How did the L1 bearing temperature and alert status change across Sebring, Salem, and East P., Ohio, and what did the NTSB report happened next?\"},\"items\":[{\"id\":\"0\",\"text\":\"## When the bearing failed\"},{\"id\":\"1\",\"text\":\"The East P., Ohio derailment began when an overheated bearing burned off the accident hopper car. Here is what you need to know about bearings, hot bearing detectors (HBDs), and what happened when the accident hopper car passed over the final three HBDs of the journey in the Ohio towns of Sebring, Salem, and East P., Ohio.\"},{\"id\":\"2\",\"text\":\"A bearing (A) fits around the end of a railcar axle. The part of the axle extending into the bearing is called the journal (B) . The bearing's roller components allow the journal to turn within the bearing with little friction. However, when a bearing fails, the resulting friction can be catastrophic. In East P., Ohio, the journal's cylindrical shape was worn down to a cone (C) . ( See box below .)\"},{\"id\":\"3\",\"text\":\"Overheated wheel bearings are the most common mechanical or electrical cause of rail accidents. Because overheated bearings are so dangerous, railroad tracks are equipped with HBDs, devices located along the track that measure the bearing's temperature relative to the ambient temperature (D) . Ambient temperature means the temperature outside. For example, the ambient temperature at Sebring, Ohio, was measured by a thermometer at the nearby signal bungalow.\"},{\"id\":\"4\",\"text\":\"Bearings can go from below-alarm temperature to seized in less than a minute. In this accident, however, fire was first photographed near the L1 bearing by a surveillance camera in Salem, Ohio (right), 39 minutes before the train derailed in East P., Ohio.\"},{\"id\":\"5\",\"text\":\"The bearing that caused the accident hopper car to derail was its L1 bearing, which is on the right side of the car, facing the direction of travel.\"},{\"id\":\"6\",\"text\":\"Railcar axles are numbered starting from the end of the railcar with the brake wheel (the 'B end'). When the B end is in the front of the car (when it faces the direction of travel) the 'left-side' wheels and bearings are on the right.\"},{\"id\":\"7\",\"text\":\"## The undetected hot bearing burns off\"},{\"id\":\"8\",\"text\":\"1. HBDs (D) measure the temperature on the outside of the bearing cup. When a bearing heats up inside, it may take 30-60 minutes for the HBD to sense it from the outside. The HBD in Salem measured a temperature of only 103° above ambient, despite the fact that the bearing was on fire. So, it generated an alert to the Wayside Help Desk, not an alarm to the train crew.\"},{\"id\":\"9\",\"text\":\"2. By the time the alarm sounded at the East P., Ohio HBD, the journal was worn down to a cone, and it was too late. The hopper car and 37 others derailed as the train's emergency braking system activated.\"},{\"id\":\"10\",\"text\":\"## On fire but not 'critical'\"},{\"id\":\"11\",\"text\":\"At 7:37 p.m. , at the Sebring HBD, the reading for the L1 bearing was only 38°F above ambient -not high enough to trigger an alert. The R1 bearing measured 20°F above ambient. But the L1 bearing was hotter.\"},{\"id\":\"12\",\"text\":\"At 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted.\"},{\"id\":\"13\",\"text\":\"About 8:52 p.m ., the train went over the East P., Ohio HBD. Less than a minute later, the HBD recorded a temperature of 253°F above ambient at the L1 bearing; bearing R1 remained at 20°F above ambient. The HBD immediately transmitted a critical alarm, which was broadcast in the locomotive cab.\"},{\"id\":\"14\",\"text\":\"Under NS rules, if a train's crew hears a critical alarm, they must stop the train right away and inspect the wheelset that triggered the alarm. The engineer began to slow the train before 8:54 p.m.\"},{\"id\":\"15\",\"text\":\"But it was too late; the hopper car and 37 others derailed as the train's emergency braking system activated.\"},{\"id\":\"16\",\"text\":\"The Wayside Help Desk in Atlanta, Georgia, monitors alerts from NS's HBDs. Critical alarms broadcast straight to the train crew. For noncritical alerts, the analyst monitors the train over HBDs it crosses later, paying attention to temperature trends. The single analyst on duty did not notice the Salem noncritical alert, as he dealt with higher-priority alerts. But if he did notice the alert, that would not have prevented the derailment. The analyst said that if he had seen the Salem HBD alert in real time, he would have followed NS rules and watched for alerts from HBDs it went over later. The next HBD the accident train went over was in East P., Ohio.\"},{\"id\":\"17\",\"text\":\"In the East P., Ohio derailment, the alarm was received by the lead locomotive radio and broadcast over the in-cab speakers: ' Critical alarm, critical alarm, critical alarm. Norfolk Southern milepost 49.8, track 1 hotbox, axle 101, south rail. '\"}]}"
Build the "Rerank" capability into my app using the Superlinked Inference Engine (SIE).

Context
- SIE is an OpenAI-style inference API. Python SDK: `from sie_sdk import SIEClient`; TypeScript: `@superlinked/sie-sdk`.
- Base URL: https://api.superlinked.com (or my regional endpoint). Auth: Bearer key from env `SIE_API_KEY` (never hard-code it).
- Model: Qwen/Qwen3-Reranker-4B (SIE primitive: /score). Keep the model id configurable.

Task
- Input: a query string plus a list of candidate documents.
- Behaviour: reorder the candidates by true relevance to the query using the cross-encoder
- Call the selected SIE primitive once per request and map the response into your domain type.

Deliverables
- A typed client wrapper, an application-level function for this task, error handling for timeouts/empty input, and unit tests with a stubbed client.
- Wire it into my existing stack (ask me which framework if unclear) and add a short usage example.
Output

On fire but not 'critical'

Sebring · 7:37 p.m.

The L1 bearing measured 38°F above ambient, below the alert threshold.

Salem · 8:13 p.m.

The reading reached 103°F above ambient. A surveillance camera showed fire near the bearing.

Image region preserved

East P., Ohio · 8:52 p.m.

The detector recorded 253°F above ambient and transmitted a critical alarm to the locomotive cab.

  • Extracts clean, agent-ready markdown
  • Keeps tables and layout intact
Output

At 8:13 p.m.event time , at the Salem HBDlocation, the reading for the L1 bearingbearing was 103°Fdegrees above ambient above ambient . This reading was high enough to trigger a noncriticalalert status alert to the Wayside Help Deskalert recipient ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearingcamera observation ( right ). Surveillance cameras in Sebringlocation also captured images of the train, and no anomalies were noted.

8:13 p.m.event timeSalem HBDlocationL1 bearingbearing103°Fdegrees above ambientnoncriticalalert statusWayside Help Deskalert recipientSalemlocationfire near the bearingcamera observationSebringlocation
9 matches across 7 types
  • Pulls typed entities straight from raw text
Output
At 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted.
0.65
At 7:37 p.m. , at the Sebring HBD, the reading for the L1 bearing was only 38°F above ambient -not high enough to trigger an alert. The R1 bearing measured 20°F above ambient. But the L1 bearing was hotter.
0.62
The East P., Ohio derailment began when an overheated bearing burned off the accident hopper car. Here is what you need to know about bearings, hot bearing detectors (HBDs), and what happened when the accident hopper car passed over the final three HBDs of the journey in the Ohio towns of Sebring, Salem, and East P., Ohio.
0.59
About 8:52 p.m ., the train went over the East P., Ohio HBD. Less than a minute later, the HBD recorded a temperature of 253°F above ambient at the L1 bearing; bearing R1 remained at 20°F above ambient. The HBD immediately transmitted a critical alarm, which was broadcast in the locomotive cab.
0.58
Bearings can go from below-alarm temperature to seized in less than a minute. In this accident, however, fire was first photographed near the L1 bearing by a surveillance camera in Salem, Ohio (right), 39 minutes before the train derailed in East P., Ohio.
0.58
The Wayside Help Desk in Atlanta, Georgia, monitors alerts from NS's HBDs. Critical alarms broadcast straight to the train crew. For noncritical alerts, the analyst monitors the train over HBDs it crosses later, paying attention to temperature trends. The single analyst on duty did not notice the Salem noncritical alert, as he dealt with higher-priority alerts. But if he did notice the alert, that would not have prevented the derailment. The analyst said that if he had seen the Salem HBD alert in real time, he would have followed NS rules and watched for alerts from HBDs it went over later. The next HBD the accident train went over was in East P., Ohio.
0.55
In the East P., Ohio derailment, the alarm was received by the lead locomotive radio and broadcast over the in-cab speakers: ' Critical alarm, critical alarm, critical alarm. Norfolk Southern milepost 49.8, track 1 hotbox, axle 101, south rail. '
0.54
1. HBDs (D) measure the temperature on the outside of the bearing cup. When a bearing heats up inside, it may take 30-60 minutes for the HBD to sense it from the outside. The HBD in Salem measured a temperature of only 103° above ambient, despite the fact that the bearing was on fire. So, it generated an alert to the Wayside Help Desk, not an alarm to the train crew.
0.49
2. By the time the alarm sounded at the East P., Ohio HBD, the journal was worn down to a cone, and it was too late. The hopper car and 37 others derailed as the train's emergency braking system activated.
0.48
Overheated wheel bearings are the most common mechanical or electrical cause of rail accidents. Because overheated bearings are so dangerous, railroad tracks are equipped with HBDs, devices located along the track that measure the bearing's temperature relative to the ambient temperature (D) . Ambient temperature means the temperature outside. For example, the ambient temperature at Sebring, Ohio, was measured by a thermometer at the nearby signal bungalow.
0.48
## The undetected hot bearing burns off
0.46
## When the bearing failed
0.41
The bearing that caused the accident hopper car to derail was its L1 bearing, which is on the right side of the car, facing the direction of travel.
0.41
Under NS rules, if a train's crew hears a critical alarm, they must stop the train right away and inspect the wheelset that triggered the alarm. The engineer began to slow the train before 8:54 p.m.
0.40
But it was too late; the hopper car and 37 others derailed as the train's emergency braking system activated.
0.40
## On fire but not 'critical'
0.37
A bearing (A) fits around the end of a railcar axle. The part of the axle extending into the bearing is called the journal (B) . The bearing's roller components allow the journal to turn within the bearing with little friction. However, when a bearing fails, the resulting friction can be catastrophic. In East P., Ohio, the journal's cylindrical shape was worn down to a cone (C) . ( See box below .)
0.36
Railcar axles are numbered starting from the end of the railcar with the brake wheel (the 'B end'). When the B end is in the front of the car (when it faces the direction of travel) the 'left-side' wheels and bearings are on the right.
0.30
Top match at 0.65 · 18 candidates ranked
  • Ranks by meaning, not keyword overlap
  • Scores every candidate against your query
Output
At 8:13 p.m. , at the Salem HBD, the reading for the L1 bearing was 103°F above ambient . This reading was high enough to trigger a noncritical alert to the Wayside Help Desk ( see box ), but not to the crew. A surveillance camera picture from Salem showed fire near the bearing ( right ). Surveillance cameras in Sebring also captured images of the train, and no anomalies were noted.
↑ up 120.95
The East P., Ohio derailment began when an overheated bearing burned off the accident hopper car. Here is what you need to know about bearings, hot bearing detectors (HBDs), and what happened when the accident hopper car passed over the final three HBDs of the journey in the Ohio towns of Sebring, Salem, and East P., Ohio.
– held0.94
At 7:37 p.m. , at the Sebring HBD, the reading for the L1 bearing was only 38°F above ambient -not high enough to trigger an alert. The R1 bearing measured 20°F above ambient. But the L1 bearing was hotter.
↑ up 90.91
Bearings can go from below-alarm temperature to seized in less than a minute. In this accident, however, fire was first photographed near the L1 bearing by a surveillance camera in Salem, Ohio (right), 39 minutes before the train derailed in East P., Ohio.
↑ up 10.91
About 8:52 p.m ., the train went over the East P., Ohio HBD. Less than a minute later, the HBD recorded a temperature of 253°F above ambient at the L1 bearing; bearing R1 remained at 20°F above ambient. The HBD immediately transmitted a critical alarm, which was broadcast in the locomotive cab.
↑ up 90.84
1. HBDs (D) measure the temperature on the outside of the bearing cup. When a bearing heats up inside, it may take 30-60 minutes for the HBD to sense it from the outside. The HBD in Salem measured a temperature of only 103° above ambient, despite the fact that the bearing was on fire. So, it generated an alert to the Wayside Help Desk, not an alarm to the train crew.
↑ up 30.47
2. By the time the alarm sounded at the East P., Ohio HBD, the journal was worn down to a cone, and it was too late. The hopper car and 37 others derailed as the train's emergency braking system activated.
↑ up 30.38
In the East P., Ohio derailment, the alarm was received by the lead locomotive radio and broadcast over the in-cab speakers: ' Critical alarm, critical alarm, critical alarm. Norfolk Southern milepost 49.8, track 1 hotbox, axle 101, south rail. '
↑ up 100.08
## The undetected hot bearing burns off
↓ down 10.03
## When the bearing failed
↓ down 90.03
The Wayside Help Desk in Atlanta, Georgia, monitors alerts from NS's HBDs. Critical alarms broadcast straight to the train crew. For noncritical alerts, the analyst monitors the train over HBDs it crosses later, paying attention to temperature trends. The single analyst on duty did not notice the Salem noncritical alert, as he dealt with higher-priority alerts. But if he did notice the alert, that would not have prevented the derailment. The analyst said that if he had seen the Salem HBD alert in real time, he would have followed NS rules and watched for alerts from HBDs it went over later. The next HBD the accident train went over was in East P., Ohio.
↑ up 60.01
Overheated wheel bearings are the most common mechanical or electrical cause of rail accidents. Because overheated bearings are so dangerous, railroad tracks are equipped with HBDs, devices located along the track that measure the bearing's temperature relative to the ambient temperature (D) . Ambient temperature means the temperature outside. For example, the ambient temperature at Sebring, Ohio, was measured by a thermometer at the nearby signal bungalow.
↓ down 80.01
The bearing that caused the accident hopper car to derail was its L1 bearing, which is on the right side of the car, facing the direction of travel.
↓ down 70.01
## On fire but not 'critical'
↓ down 30.01
But it was too late; the hopper car and 37 others derailed as the train's emergency braking system activated.
↑ up 10.01
A bearing (A) fits around the end of a railcar axle. The part of the axle extending into the bearing is called the journal (B) . The bearing's roller components allow the journal to turn within the bearing with little friction. However, when a bearing fails, the resulting friction can be catastrophic. In East P., Ohio, the journal's cylindrical shape was worn down to a cone (C) . ( See box below .)
↓ down 130.00
Under NS rules, if a train's crew hears a critical alarm, they must stop the train right away and inspect the wheelset that triggered the alarm. The engineer began to slow the train before 8:54 p.m.
↓ down 20.00
Railcar axles are numbered starting from the end of the railcar with the brake wheel (the 'B end'). When the B end is in the front of the car (when it faces the direction of travel) the 'left-side' wheels and bearings are on the right.
↓ down 110.00
  • Reorders by true relevance
  • Scored by meaning, keyword-free

Compare models for this task

PRICE
$ / 1M input tokens
$0.02 OpenAI 3-small
$0.02 Voyage voyage-4-lite
$0.0377 SIE Arctic Embed L v2
Voyage voyage-4-large $0.12
SIE Qwen3 Embedding 4B $0.13
OpenAI 3-large $0.13
QUALITY
nDCG@10 · NFCorpus
SIE Qwen3 Embedding 4B 0.41
0.40
0.39
SIE Arctic Embed L v2 0.35
0.34
0.33
LATENCY
p50 ms
36ms SIE Arctic Embed L v2
44ms
50ms
SIE Qwen3 Embedding 4B 150ms
165ms
178ms
PRICE
$ / 1M input tokens $ / 1M pairs
$0.02 Voyage rerank-2.5-lite
$0.025 ZeroEntropy zerank-2
Voyage rerank-2.5 $0.05
Jina Reranker v3 $0.05
$6.63 SIE Qwen3 Reranker 0.6B
SIE Qwen3 Reranker 4B $61.89
QUALITY
nDCG@10 · AskUbuntu
SIE Qwen3 Reranker 4B 0.70
0.68
0.67
SIE Qwen3 Reranker 0.6B 0.65
0.64
0.63
LATENCY
p50 ms
60ms SIE Qwen3 Reranker 0.6B
75ms
150ms
SIE Qwen3 Reranker 4B 580ms
640ms
700ms
PRICE
$ / 1M input tokens
$0.0907 SIE GLiNER Multi
$0.186 SIE GLiNER2 Large
$0.2 OpenAI GPT-5.4 nano
$0.3 Google Gemini 3.5 Flash-Lite
OpenAI GPT-5.4 mini $0.75
Anthropic Claude Haiku 4.5 $1
QUALITY
F1 · CoNLL-03
SIE GLiNER Multi 0.60
0.58
0.57
SIE GLiNER2 Large 0.54
0.52
0.51
LATENCY
p50 ms
SIE GLiNER Multi 82ms
95ms
108ms
SIE GLiNER2 Large 130ms
145ms
158ms
PRICE
$ / 1k pages
$0.397 SIE docling
$1.5 AWS Textract
$1.5 Google Enterprise Document OCR
$1.5 Azure Document Intelligence Read
Mistral OCR 4 $4
QUALITY
olmOCR
SIE docling 0.32
0.30
0.30
0.29
0.29
LATENCY
p50 ms
SIE docling 214ms
240ms
260ms
270ms
300ms

Our example agent reconstructs the detector sequence from the NTSB record

View on GitHub

Detector chronology

  1. Sebring · 38°F Below alert threshold
  2. Salem · 103°F Noncritical · Help Desk
  3. East P., Ohio · 253°F Critical · locomotive cab

Five model stages collect and rank the evidence

Read the illustrated digest Docling Doc to Markdown
Extract detector events GLiNER multi v2.1 Named entities
Retrieve alert passages BGE-M3 Search
Prioritize the detector sequence Qwen3-Reranker-4B Rerank
Validate exact source spans GLiNER2 large v1 Named entities
103°F Salem reading above ambient
253°F East P., Ohio reading
39 min between final readings
38 railcars derailed

Deploy your way

Self-host with K8s

Easy & scalable deployment in your own cloud.

  • Terraform to your cloud in minutes
  • Apache-2.0, same engine as Cloud
  • Scales to zero, no bill between jobs
  • Per-tenant pools, no noisy neighbors
Agent prompt
Deploy SIE to our AWS account with the superlinked/sie/aws Terraform module. Docs: superlinked.com/docs/deploymentDeploy SIE to our GCP project with the superlinked/sie/google Terraform module. Docs: superlinked.com/docs/deploymentDeploy SIE to our Azure AKS cluster via helm install. Requirements: superlinked.com/docs/deployment
Deploy guide

Run locally

Run the same models on your own machine.

  • Runs on NVIDIA GPU or Apple Silicon
  • One command, no Docker or cluster
  • All 100+ Cloud models, fully offline
  • Same SDK and IDs, no code changes
pip install "sie-server[local]" && sie-server servepip install "sie-server[local]" && sie-server serve --device cuda
Quickstart

Contact us

Tell us about your use case and we'll get back to you shortly.

Apply for an inference grant

Free capacity on our hosted cluster for selected projects. Tell us what you run and we reply by email.