Gemma 4 31B
An open vision model that reads documents and app screens into the JSON your schema asks for. google/gemma-4-31B-it
- Accuracy
MMLU-Pro, answers right · not measured
- Time per request
Not timed one request at a time
- Serverless price
$0.45 in · $1.50 out
per 1M tokens
-
Extract more exact field values than Claude Sonnet 5.5, 88% cheaper
94.3%
Exact field values · 546 documents
See the comparison on Doc field extraction$2.12
Cost per 1,000 documents
-
Read app screenshots 69% cheaper than GPT‑6.1 Sol
94.72%
Field agreement · 100 screenshots
See the comparison on Screenshot mining$1.88
Cost · per 1,000 screenshots
Call it
from pathlib import Pathimport base64import jsonfrom sie_sdk import SIEClientclient = SIEClient(api_key="API keysk-sie-…")schema = json.loads("""schema{ "type": "object", "properties": { "items": {"anyOf": [{"type": "array", "items": {"type": "object", "properties": {···"brand": ···, "quantity": ···, "reference": ···, "productSpecsJson": {···"watts": ···, "efficiency": ···}, {"type": "null"}]}, "header": {"type": "object", "properties": {···"date": ···} }, "required": ["items", "header"], "additionalProperties": false}""")instruction = "instruction···Copy values as they appear on the document.···"prompt = instructionimage = base64.b64encode(Path(fileJPGdelivery-note.jpgbrowse).read_bytes()).decode("ascii")# Messages: the prompt as the system message, the image as the user message. Copy includes them.result = client.chat_completions( "modelgoogle/gemma-4-31B-it:h100-96k-hires-out8k-no-spec", messages, max_completion_tokens=8192, extra_body={"temperature":0,"presence_penalty":0}, # Strict JSON-schema response_format. Copy includes it.)# Copy also reads and parses the JSON reply.import { readFile } from 'node:fs/promises';
import { SIEClient } from '@superlinked/sie-sdk';
const client = new SIEClient(undefined, { apiKey: 'sk-sie-…' });
const schema = {
"type": "object",
"properties": {
"items": {"anyOf": [{"type": "array", "items": {"type": "object", "properties": {···"brand": ···, "quantity": ···, "reference": ···, "productSpecsJson": {···"watts": ···, "efficiency": ···}, {"type": "null"}]},
"header": {"type": "object", "properties": {···"date": ···}
},
"required": ["items", "header"],
"additionalProperties": false
};
const instruction = "···Copy values as they appear on the document.···";
const prompt = instruction;
const image = await readFile("delivery-note.jpg");
const result = await client.chatCompletions({
model: 'google/gemma-4-31B-it:h100-96k-hires-out8k-no-spec',
// Messages: the prompt as the system message, the image as the user message. Copy includes them.
max_completion_tokens: 8192,
temperature: 0,
presence_penalty: 0,
// Strict JSON-schema response_format. Copy includes it.
});
// Copy also reads and parses the JSON reply.document_image=$(base64 < 'delivery-note.jpg' | tr -d '\n')
curl https://api.superlinked.com/v1/chat/completions \
-H "Authorization: Bearer sk-sie-…" \
-H "Content-Type: application/json" \
-d "{\"model\":\"google/gemma-4-31B-it:h100-96k-hires-out8k-no-spec\",\"messages\":[{\"role\":\"system\",\"content\":\"···Copy values as they appear on the document.···\"},{\"role\":\"user\",\"content\":[{\"type\":\"image_url\",\"image_url\":{\"url\":\"data:image/jpeg;base64,$document_image\"}}]}],\"max_completion_tokens\":8192,\"temperature\":0,\"presence_penalty\":0,\"response_format\":{\"type\":\"json_schema\",\"json_schema\":{\"name\":\"fields\",\"strict\":true,\"schema\":{\"type\":\"object\",\"properties\":{\"items\":{\"anyOf\":[{\"type\":\"array\",\"items\":{\"type\":\"object\",\"properties\":{\"sn\":{\"anyOf\":[{\"type\":\"number\"},{\"type\":\"null\"}]},\"type\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"LED Luminaire\",\"Power Supply\",\"Control System\",\"Mounting Bracket\",\"LED Driver\",\"Cable Assembly\",\"Junction Box\",\"Sensor Module\"]},{\"type\":\"null\"}]},\"unit\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"pcs\",\"set\",\"unit\"]},{\"type\":\"null\"}]},\"brand\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"Philips\",\"Osram\",\"Schneider Electric\",\"Siemens\",\"ABB\",\"Lutron\",\"Legrand\",\"Mean Well\"]},{\"type\":\"null\"}]},\"quantity\":{\"anyOf\":[{\"type\":\"number\"},{\"type\":\"null\"}]},\"reference\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"productSpecsJson\":{\"type\":\"object\",\"properties\":{\"power\":{\"type\":\"object\",\"properties\":{\"watts\":{\"anyOf\":[{\"type\":\"number\"},{\"type\":\"null\"}]},\"efficiency\":{\"anyOf\":[{\"description\":\"Efficiency in percentage, 0-100\",\"type\":\"number\"},{\"type\":\"null\"}]},\"powerFactor\":{\"anyOf\":[{\"type\":\"number\"},{\"type\":\"null\"}]},\"inputVoltage\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"100-240v\",\"100-277v\",\"220-240v\"]},{\"type\":\"null\"}]}},\"required\":[\"watts\",\"efficiency\",\"powerFactor\",\"inputVoltage\"],\"additionalProperties\":false},\"ratings\":{\"type\":\"object\",\"properties\":{\"protection\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"ip65\",\"ip66\",\"ip67\"]},{\"type\":\"null\"}]}},\"required\":[\"protection\"],\"additionalProperties\":false},\"physical\":{\"type\":\"object\",\"properties\":{\"weight\":{\"anyOf\":[{\"type\":\"number\"},{\"type\":\"null\"}]},\"material\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"aluminum\",\"steel\",\"plastic\"]},{\"type\":\"null\"}]},\"mountingType\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"surface\",\"recessed\",\"track\"]},{\"type\":\"null\"}]}},\"required\":[\"weight\",\"material\",\"mountingType\"],\"additionalProperties\":false}},\"required\":[\"power\",\"ratings\",\"physical\"],\"additionalProperties\":false}},\"required\":[\"sn\",\"type\",\"unit\",\"brand\",\"quantity\",\"reference\",\"productSpecsJson\"],\"additionalProperties\":false}},{\"type\":\"null\"}]},\"header\":{\"type\":\"object\",\"properties\":{\"to\":{\"type\":\"object\",\"properties\":{\"tel\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"email\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"location\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"attention\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"reference\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"vatNumber\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"companyName\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]}},\"required\":[\"tel\",\"email\",\"location\",\"attention\",\"reference\",\"vatNumber\",\"companyName\"],\"additionalProperties\":false},\"date\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"from\":{\"type\":\"object\",\"properties\":{\"tel\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"email\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"location\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"reference\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"vatNumber\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"companyName\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]}},\"required\":[\"tel\",\"email\",\"location\",\"reference\",\"vatNumber\",\"companyName\"],\"additionalProperties\":false},\"projectName\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]}},\"required\":[\"to\",\"date\",\"from\",\"projectName\"],\"additionalProperties\":false}},\"required\":[\"items\",\"header\"],\"additionalProperties\":false}}}}"Build the "Doc field extraction" 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: google/gemma-4-31B-it:h100-96k-hires-out8k-no-spec (OpenAI-compatible endpoint: /v1/chat/completions). Keep the model id configurable.
Task
- Input: an uploaded document plus the JSON schema for its fields.
- Behaviour: parse the document, then return fields matching the JSON schema
- Send one POST /v1/chat/completions request (SDK: chat_completions / chatCompletions) with a system message and a user message. The system message is "Extract data from the document image into the JSON schema. Copy values as they appear on the document. Use null for a field the document does not contain. Return only the JSON.\n\nJSON schema:\n{\n \"type\": \"object\",\n \"required\": [\n \"header\",\n \"items\"\n ],\n \"properties\": {\n \"items\": {\n \"type\": \"array\",\n \"items\": {\n \"type\": \"object\",\n \"required\": [\n \"sn\",\n \"type\",\n \"brand\",\n \"reference\",\n \"productSpecsJson\",\n \"unit\",\n \"quantity\"\n ],\n \"properties\": {\n \"sn\": {\n \"type\": \"number\"\n },\n \"type\": {\n \"enum\": [\n \"LED Luminaire\",\n \"Power Supply\",\n \"Control System\",\n \"Mounting Bracket\",\n \"LED Driver\",\n \"Cable Assembly\",\n \"Junction Box\",\n \"Sensor Module\"\n ],\n \"type\": \"string\"\n },\n \"unit\": {\n \"enum\": [\n \"pcs\",\n \"set\",\n \"unit\"\n ],\n \"type\": \"string\"\n },\n \"brand\": {\n \"enum\": [\n \"Philips\",\n \"Osram\",\n \"Schneider Electric\",\n \"Siemens\",\n \"ABB\",\n \"Lutron\",\n \"Legrand\",\n \"Mean Well\"\n ],\n \"type\": \"string\"\n },\n \"quantity\": {\n \"type\": \"number\"\n },\n \"reference\": {\n \"type\": \"string\"\n },\n \"productSpecsJson\": {\n \"type\": \"object\",\n \"required\": [\n \"power\",\n \"physical\",\n \"ratings\"\n ],\n \"properties\": {\n \"power\": {\n \"type\": \"object\",\n \"required\": [\n \"watts\",\n \"inputVoltage\"\n ],\n \"properties\": {\n \"watts\": {\n \"type\": \"number\"\n },\n \"efficiency\": {\n \"type\": \"number\",\n \"description\": \"Efficiency in percentage, 0-100\"\n },\n \"powerFactor\": {\n \"type\": \"number\"\n },\n \"inputVoltage\": {\n \"enum\": [\n \"100-240v\",\n \"100-277v\",\n \"220-240v\"\n ],\n \"type\": \"string\"\n }\n }\n },\n \"ratings\": {\n \"type\": \"object\",\n \"required\": [\n \"protection\"\n ],\n \"properties\": {\n \"protection\": {\n \"enum\": [\n \"ip65\",\n \"ip66\",\n \"ip67\"\n ],\n \"type\": \"string\"\n }\n }\n },\n \"physical\": {\n \"type\": \"object\",\n \"required\": [\n \"material\",\n \"mountingType\"\n ],\n \"properties\": {\n \"weight\": {\n \"type\": \"number\"\n },\n \"material\": {\n \"enum\": [\n \"aluminum\",\n \"steel\",\n \"plastic\"\n ],\n \"type\": \"string\"\n },\n \"mountingType\": {\n \"enum\": [\n \"surface\",\n \"recessed\",\n \"track\"\n ],\n \"type\": \"string\"\n }\n }\n }\n }\n }\n }\n }\n },\n \"header\": {\n \"type\": \"object\",\n \"required\": [\n \"from\",\n \"to\",\n \"date\",\n \"projectName\"\n ],\n \"properties\": {\n \"to\": {\n \"type\": \"object\",\n \"required\": [\n \"companyName\",\n \"reference\",\n \"location\",\n \"attention\",\n \"tel\",\n \"email\",\n \"vatNumber\"\n ],\n \"properties\": {\n \"tel\": {\n \"type\": \"string\"\n },\n \"email\": {\n \"type\": \"string\"\n },\n \"location\": {\n \"type\": \"string\"\n },\n \"attention\": {\n \"type\": \"string\"\n },\n \"reference\": {\n \"type\": \"string\"\n },\n \"vatNumber\": {\n \"type\": \"string\"\n },\n \"companyName\": {\n \"type\": \"string\"\n }\n }\n },\n \"date\": {\n \"type\": \"string\"\n },\n \"from\": {\n \"type\": \"object\",\n \"required\": [\n \"companyName\",\n \"reference\",\n \"location\",\n \"tel\",\n \"email\",\n \"vatNumber\"\n ],\n \"properties\": {\n \"tel\": {\n \"type\": \"string\"\n },\n \"email\": {\n \"type\": \"string\"\n },\n \"location\": {\n \"type\": \"string\"\n },\n \"reference\": {\n \"type\": \"string\"\n },\n \"vatNumber\": {\n \"type\": \"string\"\n },\n \"companyName\": {\n \"type\": \"string\"\n }\n }\n },\n \"projectName\": {\n \"type\": \"string\"\n }\n }\n }\n }\n}" verbatim, without appending the response schema. The user message holds only an image_url part. Base64-encode the original local image bytes as a canonical data:image/png, data:image/jpeg or data:image/webp URI (normalize jpg to image/jpeg). Use max_completion_tokens 8192, temperature 0 and response_format {type: "json_schema", json_schema: {name: "fields", strict: true, schema}} with the submitted schema unchanged. Parse choices[0].message.content as JSON and fail loudly when it is missing or invalid.
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. Edit the code and run it live Sign up to run this example.
Get started no credit card required
Output
header.date 2025-02-12
| reference | brand | quantity | watts | efficiency |
|---|---|---|---|---|
| ASDZPDI5YL | Lutron | 50 | 103 | 95 |
| FL5GMWQOUQ | Osram | 5 | 188 | 93 |
Next to the other language models
| Model | MMLU-Pro, answers right | Time per request | Serverless price | Size |
|---|---|---|---|---|
| Gemma 4 31B google | Not measured | Not timed | $0.45 in · $1.50 out per 1M tokens | 31.3B |
| Qwen3.8 27B Qwen | Not measured | 0.74 s per complete reply | $0.25 in · $2.75 out per 1M tokens | 27.8B |
| GLM-5.3-Flash zai-org | Not measured | Not timed | $0.15 in · $0.50 out per 1M tokens | 321.3B |
| Qwen3.5 4B Qwen | 59.3% | Not timed | $0.305 in · $1.53 out per 1M tokens | 4.7B |
| Qwen3 4B Instruct 2507 Qwen | 53.3% | Not timed | Run in your cloud | 4.0B |
Facts
- Size
- 31.3B parameters
- License
- Apache-2.0
- Reads
- Text (up to 8,192 tokens), image
- Endpoint
-
generate
Evaluation
- Accuracy
- Run on SIE against the public dataset named in each figure
- Time
- SIE timed on a GPU server with the client alongside, one request at a time
- Prices
- Serverless list prices
- Sources
- Task page runs and their inputs
Run Gemma 4 31B on Serverless or in your own cloud
SIE Serverless
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
SIE Cluster
Kubernetes in your own 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. SIE Local
The open-source server on your own machine.
- NVIDIA GPU, 80 GB
- 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-cuda13-sglang-cu130