Rank document pages 4% better than Voyage multimodal‑3.5

Tomoro ColQwen3 searches PDFs, slides and scans directly from their page images, including tables and charts. SIE is the open-source inference engine that serves it through one API.

Get started Run in your cloud

SIE Tomoro ColQwen3 4B

Right page

QUERY

What is the 2026 SEMIMONTHLY standard withholding for Married Filing Jointly with adjusted wages of $1,000?

1 · Page 20IRS Publication 15-T · 2026

SEMIMONTHLY Payroll Period

Table excerpt

Wage bracket $995 to $1,010; Married Filing Jointly standard withholding $0

Excerpt from the first of 8 ranked pages. Recorded query

A similar-looking page can send your answer to the wrong source

Voyage multimodal-3.5

Wrong page first

Question

FDA enhancements to REMS programs 2017 2018 efforts

Open full source page ↗

Approved drugs, with no account of the requested program enhancements. The answer page came 3rd.

FDA, Office of Surveillance and Epidemiology, 2017 update

OCR + OpenAI text-embedding-3-large

Wrong page first

Question

What was Romania’s employment rate for people aged 15 to 24 in 2020?

Open full source page ↗

No figure for 2020 on this page. The answer page came 76th.

European Commission, Labour market and wage developments in Europe 2024 (CC BY 4.0)

arXiv:2601.08620 Visual retrievers outperform text retrievers on complex documents
ViDoRe V3: A Comprehensive Evaluation of Retrieval Augmented Generation in Complex Real-World Scenarios

António Loison, Quentin Macé, Antoine Edy et al.

“current models still struggle with non-textual elements, open-ended queries, and fine-grained visual grounding”
21 April 2026
arXiv:2412.02592 ICCV 2025 None of the OCR tools tested produced text good enough for a reliable RAG index
OCR Hinders RAG: Evaluating the Cascading Impact of OCR on Retrieval-Augmented Generation

Junyuan Zhang, Qintong Zhang, Bin Wang et al.

“none is competent for constructing high-quality knowledge bases for RAG systems”
3 December 2024

Search page images 72% cheaper than Cohere Embed v5 Pro

Run this eval

Answer-page ranking

nDCG@10, of 100 ↑

Cohere Embed v5 Pro

Cohere Embed v5 Pro 64.8

SIE Tomoro ColQwen3 4B

SIE Tomoro ColQwen3 4B 64.0

Voyage multimodal-3.5

Voyage multimodal-3.5 61.0

OCR + OpenAI embeddings

OCR + OpenAI embeddings 56.0

Cost per 1,000 pages

indexing, USD ↓

SIE Tomoro ColQwen3 4B

$0.20 SIE Tomoro ColQwen3 4B

Cohere Embed v5 Pro

$0.74 Cohere Embed v5 Pro

Voyage multimodal-3.5

Voyage multimodal-3.5 $1.11

OCR + OpenAI embeddings

OCR + OpenAI embeddings $1.57

Evaluation

Tested: ViDoRe v3: 1,816 questions across 11,624 public document pages
Scored: Human relevance grades; nDCG@10 averaged over six document sets
Prices: SIE Serverless; providers at their scored page size

Encode the page once. Search it by text.

from pathlib import Path
from urllib.request import urlopen
from sie_sdk import SIEClient
client = SIEClient(api_key="API keysk-sie-…")
model = "modelTomoroAI/tomoro-colqwen3-embed-4b:compact"
query = "queryWhat is the 2026 SEMIMONTHLY standard withholding for Married Filing Jointly with adjusted wages of $1,000?"
pages = [
    "irs-p15t-page-13.png", "irs-p15t-page-14.png",
    # Copy and Run include all 8 pages.
]
# Pinned image-loading setup omitted; Copy includes it.
q = client.encode(
    model, [{"text": query}],
    output_types=["multivector"], is_query=True,
)[0]["multivector"]
docs = client.encode(
    model, [{"images": [image]} for image in images],
    output_types=["multivector"], is_query=False,
)
print(q)
print(docs)

import { readFile } from 'node:fs/promises';
import { SIEClient } from '@superlinked/sie-sdk';

const client = new SIEClient(undefined, { apiKey: 'sk-sie-…' });

const model = 'TomoroAI/tomoro-colqwen3-embed-4b:compact';
const query = "What is the 2026 SEMIMONTHLY standard withholding for Married Filing Jointly with adjusted wages of $1,000?";
const pages = [
    "irs-p15t-page-13.png", "irs-p15t-page-14.png",
    // Copy and Run include all 8 pages.
];
// Pinned image-loading setup omitted; Copy includes it.

const [q] = await client.encode(
  model, [{ text: query }],
  { outputTypes: ['multivector'], isQuery: true },
);
const docs = await client.encode(
  model, images.map((image) => ({ images: [image] })),
  { outputTypes: ['multivector'], isQuery: false },
);
console.log(q, docs);

page_0=$(curl --fail --silent --show-error --location 'https://huggingface.co/datasets/superlinked/sie-task-evidence/resolve/74f4e0eb2676a95b625783932796edac12040644/task-visual-examples-2026/irs-p15t-page-13.png' | base64 | tr -d '\n')
page_1=$(curl --fail --silent --show-error --location 'https://huggingface.co/datasets/superlinked/sie-task-evidence/resolve/74f4e0eb2676a95b625783932796edac12040644/task-visual-examples-2026/irs-p15t-page-14.png' | base64 | tr -d '\n')
page_2=$(curl --fail --silent --show-error --location 'https://huggingface.co/datasets/superlinked/sie-task-evidence/resolve/74f4e0eb2676a95b625783932796edac12040644/task-visual-examples-2026/irs-p15t-page-15.png' | base64 | tr -d '\n')
page_3=$(curl --fail --silent --show-error --location 'https://huggingface.co/datasets/superlinked/sie-task-evidence/resolve/74f4e0eb2676a95b625783932796edac12040644/task-visual-examples-2026/irs-p15t-page-16.png' | base64 | tr -d '\n')
page_4=$(curl --fail --silent --show-error --location 'https://huggingface.co/datasets/superlinked/sie-task-evidence/resolve/74f4e0eb2676a95b625783932796edac12040644/task-visual-examples-2026/irs-p15t-page-17.png' | base64 | tr -d '\n')
page_5=$(curl --fail --silent --show-error --location 'https://huggingface.co/datasets/superlinked/sie-task-evidence/resolve/74f4e0eb2676a95b625783932796edac12040644/task-visual-examples-2026/irs-p15t-page-18.png' | base64 | tr -d '\n')
page_6=$(curl --fail --silent --show-error --location 'https://huggingface.co/datasets/superlinked/sie-task-evidence/resolve/74f4e0eb2676a95b625783932796edac12040644/task-visual-examples-2026/irs-p15t-page-19.png' | base64 | tr -d '\n')
page_7=$(curl --fail --silent --show-error --location 'https://huggingface.co/datasets/superlinked/sie-task-evidence/resolve/74f4e0eb2676a95b625783932796edac12040644/task-visual-examples-2026/irs-p15t-page-20.png' | base64 | tr -d '\n')
curl https://api.superlinked.com/v1/encode/TomoroAI%2Ftomoro-colqwen3-embed-4b%3Acompact \
  -H "Authorization: Bearer sk-sie-…" \
  -H "Content-Type: application/json" \
  -d "{\"items\":[{\"text\":\"What is the 2026 SEMIMONTHLY standard withholding for Married Filing Jointly with adjusted wages of \$1,000?\"}],\"params\":{\"output_types\":[\"multivector\"],\"options\":{\"is_query\":true}}}"

curl https://api.superlinked.com/v1/encode/TomoroAI%2Ftomoro-colqwen3-embed-4b%3Acompact \
  -H "Authorization: Bearer sk-sie-…" \
  -H "Content-Type: application/json" \
  -d "{\"items\":[{\"images\":[{\"data\":\"$page_0\",\"format\":\"png\"}]},{\"images\":[{\"data\":\"$page_1\",\"format\":\"png\"}]},{\"images\":[{\"data\":\"$page_2\",\"format\":\"png\"}]},{\"images\":[{\"data\":\"$page_3\",\"format\":\"png\"}]},{\"images\":[{\"data\":\"$page_4\",\"format\":\"png\"}]},{\"images\":[{\"data\":\"$page_5\",\"format\":\"png\"}]},{\"images\":[{\"data\":\"$page_6\",\"format\":\"png\"}]},{\"images\":[{\"data\":\"$page_7\",\"format\":\"png\"}]}],\"params\":{\"output_types\":[\"multivector\"],\"options\":{\"is_query\":false}}}"

Build the "Visual document 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: TomoroAI/tomoro-colqwen3-embed-4b:compact (SIE primitive: /encode). Keep the model id configurable.

Task
- Input: a text query plus one or more document page images.
- Behaviour: return native token vectors for the query and document page images
- Preserve the original page image bytes. Encode the text query with output_types ["multivector"] and is_query true, then all page images in one batch with output_types ["multivector"] and is_query false. Python: client.encode with output_types and is_query. TypeScript: client.encode with outputTypes and isQuery. Return the native token vectors without application scoring or ranking.

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.

Run

Output

Recorded vectors, ranked by MaxSimTop 5 of 8

Table header with Married Filing Jointly and Standard withholding columns — 1 · Page 2028.177 MaxSim

2
Page 1726.029 MaxSim
3
Page 1825.723 MaxSim
4
Page 1925.561 MaxSim
5
Page 1525.544 MaxSim

Encode returns token vectors. These recorded vectors are ranked here with MaxSim.

Index a million PDF pages for $200

Serverless API

We host the models; you call one API.

No idle GPUs, pay for what you use
Fits your stack: SDK, API, CLI, MCP
Self-host the same open-source engine
SOC 2 Type 2, US or EU data residency
Create an account to try Serverless

Get started

Self-hosted (Kubernetes)

Your data stays in your cloud, backed by our engineers.

Terraform modules for your cloud
Apache-2.0, same engine as Serverless
GPU workers scale to zero between jobs
Separate worker pools for each tenant

Agent prompt Deploy SIE to our AWS account with the superlinked/sie/aws Terraform module. Docs: superlinked.com/docs/deployment.

Deploy SIE to our Google Cloud project with the superlinked/sie/google Terraform module. Docs: superlinked.com/docs/deployment.

Deploy SIE to our Azure subscription with the superlinked/sie/azure Terraform module. Docs: superlinked.com/docs/deployment.

Deploy SIE to our Alibaba Cloud account with the superlinked/sie/alicloud Terraform module. Docs: superlinked.com/docs/deployment.

Talk to us Deploy guide

Run locally

The open-source server on your own machine.

NVIDIA
One container, no cluster
Process your data fully offline
Same SDK and model IDs as Serverless

docker run --gpus all -p 8080:8080 ghcr.io/superlinked/sie-server:latest-cuda12-default

Quickstart

Switch to SIE in 5 minutes

import voyageai
from PIL import Image

client = voyageai.Client()
model = "voyage-multimodal-3.5"
page = client.multimodal_embed(
    [[Image.open("page.png")]], model=model, input_type="document",
).embeddings[0]
query = client.multimodal_embed(
    [[question]], model=model, input_type="query",
).embeddings[0]

SIE

from pathlib import Path
from sie_sdk import SIEClient

client = SIEClient(api_key="sk-sie-...")
model = "TomoroAI/tomoro-colqwen3-embed-4b:compact"
page = client.encode(
    model, [{"images": [Path("page.png").read_bytes()]}], is_query=False,
)[0]["multivector"]
query = client.encode(model, [{"text": question}], is_query=True)[0]["multivector"]