Support
Route tickets to the right queue
Map a request to the team that can handle it.
Invoice address change → billing | technical | account
EU tier · GDPR data handling · EU
Fast decisions.
Hosted in the EU.
Route, classify and score with an API. EU-tier requests run on contracted operators in the EU; the DPA is accepted at signup, and prompts and answers are not stored.
1 JevBench estimates for the same decision tasks versus a 27B reasoning LLM; not a measurement of the hosted service. Method and context →
Request → responseIllustrative samples
REQUEST / choice
state: Mia owns a red bicycle.
question: Which color is the bicycle?
ANSWER / probabilities
REQUEST / choice
state: a business customer asks to update an invoice address.
question: Which support queue fits?
ANSWER / probabilities
REQUEST / noul
state: a viewer deletes a workspace.
question: Is this action allowed?
ANSWER / P(true)
Germancompany
Hosted in the EUEU
Built forGDPR
Aligned with theEU AI Act
DPA (AVV)included
Encryptedin transit
ISO/IEC 27001data centres
Certifications held by our hosting providers.
Practical decisions
Routine decisions with EU processing.
Send context and a named question; use the typed answer in your workflow.
Product / Safety
Review actions against policy
Check a proposed action. Keep deterministic access controls in your application.
Viewer deletes workspace → allowed?
Quality
Score drafts against a rubric
Grade replies or summaries and escalate uncertain cases.
Draft + rubric → score
Ops / Docs
Send documents to the right workflow
Classify document text before handing it off.
Document → invoice | contract | other
Illustrative workflows. Validate representative cases on your own data before automating decisions.
Data flow
Where your data goes
Every EU request stays in the EU from the first byte to the last.
- Your appClient
- TLS 1.3
- API gateway · EUEU
- WireGuard
- GPU inference · EUEU
Answer returns over the same encrypted path
Two approaches
EU processing and Peer-to-Peer capacity
EU-tier requests always stay in the EU and have priority. Allowing worldwide Peer-to-Peer processing for a key does not change that.
Peer-to-Peer starts on spare EU capacity. When it is busy, Peer-to-Peer may use on-demand GPU capacity rented through Lium (lium.io). Independent third-party providers operate the GPU hardware in various countries, including outside the EU/EEA. This happens only for keys with “Allow worldwide processing” explicitly enabled. Without opt-in, Peer-to-Peer stays in the EU and may be queued or rejected when busy. Existing customers keep EU-only processing unless they opt in for the individual key. Request and response content is not stored on any node.
Contracts
What you get in writing
DPA
Data Processing Agreement under Art. 28 GDPR, accepted at signup.
Read the DPA →Sub-processors
Every provider with purpose and location.
See the list →Privacy policy
Which account data we process and why.
Read privacy policy →Impressum
German company, reachable contacts.
Open imprint →Security overview
Encryption, access and operations at a glance.
Read security overview →The numbers, with context
Less time. Less cost. Lower peak accuracy.
Use a decision model for narrow tasks where it meets your accuracy needs. Keep an LLM for difficult reasoning and open-ended writing.
PER 1,000 DECISIONS
~40× lower estimated cost
About $0.05 versus $2.18 for a 27B reasoning LLM.
SAME DECISION TASKS
~18× faster adjusted median
About 0.35 s versus 6.49 s.
BENCHMARK RESULTS
Lower peak accuracy
The compared decision-model implementation scores below frontier LLMs on accuracy.
JevBench cost and adjusted median-latency figures are estimates for model implementations, not measurements of the hosted Decision Models service. JevBench is run by the Decision Models founder. Use a frontier LLM for open-ended writing or difficult reasoning. Method and sources → · Source data
Press context
What TechCrunch wrote about Jev
“…LLMs as we know them aren't the right solution for a lot of software because they are comparatively slow and expensive.”
EU AI Act
Aligned with the EU AI Act
Decision Models hosts open-weight general-purpose models for narrow decision tasks. We publish model cards and benchmark results, label our models clearly, and document intended use. Using the API for high-risk purposes under the AI Act requires your own assessment.
Model lineup
s1-pro
Rune · EU
Calibrated text decisions, with up to three questions per call. Rune is the main EU model.
s1-fast
Plumb-4B · short text
Short text, low latency. Runs on a batched engine, so probabilities for identical requests can differ slightly.
s1-vision
Rune · EU · one image
Choices with one image attached.
s1-llm-auto-router
CPU
Classifies category, difficulty and stakes to help your router choose an LLM.
When Rune is unavailable, s1-pro text falls back to Winnow-12B Q8; on the EU tier and on Peer-to-Peer without worldwide opt-in s1-vision has no fallback and returns a retryable 503; Peer-to-Peer keys with worldwide processing opt-in may be served by on-demand worldwide GPUs running the Gemma 4 12B image implementation when capacity is busy. On Peer-to-Peer, Rune serves when spare queue/admission capacity permits, with a compatible single-question fallback; bundles are best-effort and may return retryable HTTP 503. Full model list: /models · Licence.
Capability, price and speed
Rune scores close to Jev on both public reference benchmarks. Our EU input rates are below Jev’s list price. Compare measured latency in the chart below. Surogate Rune 26B-A4B v3 is the main EU model, served as s1-pro for calibrated text decisions. The indexes below are public reference results — not hosted Decision Models scores — and JevBench is run by the founder of Decision Models. Method and source data: /benchmarks · download the data (JSON).
Decision Index
Balanced, chance-corrected index / 100 ↑ · v0.2.1
Snapshot: Decision Index v0.2.1, 2 October 2026 · see the current board
Ours – s1-pro · reference modelJev 1.13.0 (reference API)Other tested models
Scale 0–100
Source: Decision Index · data · retrieved 2 October 2026
Notes
Public default balanced index; equal weight across five areas. Rows retain the source order and use sequential display ranks, including tied scores. Jev is the reference API; other entries are tested model implementations. Reference results are not a hosted System1 measurement.
JevBench Capability Score
Mean of Intelligence and Calibration / 100 ↑ · v1.5.5 · cap-admitted
Snapshot: JevBench Capability Score v1.5.5, 2 October 2026 · see the current board
Ours – s1-pro · reference modelJev 1.13.0 (reference API)Other tested models
Scale 0–100
Source: JevBench Capability Score · data · retrieved 2 October 2026
Notes
Only models inside the cost and median latency caps (each 2× Jev 1.13.0) are ranked. Benchmark run by the founder of System1 Models. Tested reference engines; the hosted System1 engine has not been scored in this release.
Median latency, lower is better
Medium text · one question · warm HTTP/2 · median ms ↓
Decision ModelsJev 1.13.0
Scale 0–300 ms
Measured from Nuremberg, Germany, 2 October 2026. 33 rotated rounds on identical medium synthetic support tickets, one question, HTTP/2 persistent clients; first 3 rounds excluded as warm-up; 30 measured calls per target. p50=sample median; p95=nearest rank ceil(0.95*n). End-to-end successful-response latency; every failure retained. s1-pro has a lower median but a higher p95 than Jev. Small sample; results vary with workload and network.
Source: Artifact (JSON) · retrieved 2 October 2026
The measured latency comparison stands in the speed chart above — medians, p95 and success counts from the linked measurement artifact. Public index results are reference-only — not hosted Decision Models scores.
Method (EN)
Public synthetic text; real customer API end-to-end latency, successful-response quantiles; failures tracked separately; no retained customer payloads. p95 uses the nearest sample percentile.
Input price per million tokens (USD)
$0–0.05 per M input tokens
Decision ModelsJev 1.13.0
$0–0.05 per M input tokens
Launch pricing · output free · taxes excluded.
Source: Decision Models pricing · Jev 1.13.0 price · retrieved 2 October 2026
Cost per 1,000 decisions (USD)
$0–0.012 per 1,000 decisions
Decision ModelsJev 1.13.0
$0–0.012 per 1,000 decisions
Measured 3-question cost estimates. EU and Jev: 30/30 successes. Peer-to-Peer: 14/30 from Finland, 12/30 from Virginia; estimates cover successful calls only; Peer-to-Peer bundles are best-effort and may return retryable HTTP 503.
Source: Artifact (JSON) · retrieved 2 October 2026
Notes
2 October 2026. Medium text, 3 questions per call, from Finland. Mean billable input tokens × USD list rate / 3 decisions × 1,000. EU/Jev 10/10 successful medium calls; Peer-to-Peer 5/10, estimates cover successful calls only. Internal trial requests were not paid debits. Across all S/M/L cells: EU/Jev 30/30, Peer-to-Peer 14/30 in Finland and 12/30 in Virginia. Peer-to-Peer bundles are best-effort and may return retryable HTTP 503. Shared state billing reduces cost; large bundles can be slower than Jev.
EU pricing
Clear prices for the EU tier.
Per 1 million input tokens. Output tokens are free. Free allowance by invitation. Top up from EUR 1.
| Model | USD | EUR |
|---|---|---|
s1-fastPlumb-4B · Text | $0.034 | €0.030 |
s1-proSurogate Rune 26B-A4B v3 · Text | $0.040 | €0.035 |
Excluding applicable taxes. Private individuals and businesses worldwide can sign up and top up. Taxes are shown in checkout before you pay. All prices and calculator →
Common questions
Questions about EU processing
Do EU-tier requests leave the EU?
No. EU-tier requests run on contracted operators in the EU. Prompts and answers are not stored, logged or used for training. Read the DPA and the sub-processor list.
How does Peer-to-Peer differ from the EU tier?
Worldwide Peer-to-Peer processing must be explicitly enabled per key. EU-tier requests always remain in the EU.
Does the EU tier guarantee compliance for my application?
No blanket compliance promise applies to your application. You remain responsible for your application, the data you send and its lawful use.
How accurate is a decision model?
The benchmark estimate shows a substantial accuracy gap versus frontier LLMs. Validate representative cases and keep a frontier LLM or human review for difficult tasks. Method and results →