Jev vs DecisionNode: accuracy, price and features (2026)

The short answer

Pick DecisionNode when

You need images or video, number, point or box answers, batch pricing, or the input may try to steer the model: 0.764 against 0.600 on prompt injection.

Pick Jev when

Your decisions lean on world knowledge (0.820 against 0.698 on MMLU-Pro), you want official SDKs, or your requests must never be used for training by default.

On text overall

They are close: 0.816 for DecisionNode MoE and 0.803 for Jev, as the mean of 15 suites. Flash scores 0.784 at a lower price.

At a glance

TypeSafe Jev 1.13 and DecisionNode, feature by feature
Feature	TypeSafe Jev 1.13	DecisionNode
Maker	TypeSafe AI	Bynn Intelligence, Inc.
Models	Jev 1.13 (jev-1.13.0)	DecisionNode MoE (decisionnode-1.0), DecisionNode Flash (decisionnode-1.0-flash)
Endpoint	POST /v1/systemone	POST /v1/decide
Input	Text and JSON	Text, JSON, images and video frames
Answer types	choice, score, noul (yes/no)	choice, score, truth (yes/no), number, point, box
Options per choice	Up to 255	Up to 255
Levels per score	2 to 10	Up to 10
Context	64k tokens	64k tokens
Text, per 1M input tokens	$0.042	$0.042 MoE, $0.035 Flash
Images and video, per 1M tokens	Not supported	$0.09 MoE, $0.08 Flash
Batch jobs	Not in Jev's documented API	30% off on MoE, 50% off on Flash
Streaming sessions	Not in Jev's documented API	WebSocket sessions
Free tier	None listed	50 Flash and 20 MoE requests a day
Official SDKs	Python, JavaScript and TypeScript	None; plain HTTPS
Training on your requests	Never, by TypeSafe's statement	By default; you can opt out
Zero data retention	Enterprise customers	Every plan

Jev: TypeSafe's documentation. DecisionNode: decisionnode.com. Both checked October 12, 2026.

Accuracy, suite by suite

On the 15 text suites the means are 0.816 for DecisionNode MoE, 0.803 for Jev and 0.784 for DecisionNode Flash. MoE leads Jev clearly on 3 suites and trails clearly on 2; the other 10 are within noise. On the two image suites Jev has no result, because it reads text only. Every system measured is on the Jev benchmarks page.

Accuracy per suite: DecisionNode MoE, DecisionNode Flash and TypeSafe Jev 1.13, with 95% intervals
Suite	Items	DecisionNode MoE	TypeSafe Jev 1.13	DecisionNode Flash	MoE against Jev
Held-out decisions Typed decisions on tasks kept out of development.	636	0.940 0.921 to 0.959	0.948 0.929 to 0.965	0.918 0.896 to 0.940	Within noise, −0.8 pts
Public decision set A small public set of choice, score and yes/no questions.	117	0.880 0.817 to 0.939	0.855 0.784 to 0.916	0.821 0.743 to 0.884	Within noise, +2.5 pts
Transfer tasks Decisions phrased the way other decision APIs phrase them.	730	0.838 0.803 to 0.873	0.852 0.814 to 0.886	0.767 0.729 to 0.804	Within noise, −1.4 pts
Web tasks Pick the right element or action on a real web page.	1,106	0.765 0.735 to 0.794	0.698 0.660 to 0.734	0.753 0.723 to 0.784	MoE leads, +6.7 pts
Knowledge (MMLU-Pro) Hard multiple-choice questions that need world knowledge.	1,479	0.698 0.676 to 0.723	0.820 0.801 to 0.840	0.575 0.548 to 0.601	Jev leads, −12.2 pts
Short classifications One-line inputs sorted into a few labels.	178	0.921 0.882 to 0.955	0.904 0.860 to 0.944	0.882 0.826 to 0.933	Within noise, +1.7 pts
Open decision tasks A broad public mix of typed decision tasks.	5,011	0.795 0.784 to 0.806	0.798 0.787 to 0.809	0.770 0.758 to 0.781	Within noise, −0.3 pts
Typed decisions Mixed choice, score and yes/no questions in one request.	955	0.713 0.682 to 0.748	0.735 0.702 to 0.770	0.733 0.702 to 0.768	Within noise, −2.2 pts
New task families Task families the model never saw while it was built.	3,629	0.797 0.783 to 0.809	0.823 0.811 to 0.835	0.752 0.739 to 0.766	Jev leads, −2.6 pts
Familiar task families New items from task families the model knows.	4,100	0.796 0.784 to 0.809	0.774 0.763 to 0.787	0.766 0.752 to 0.778	Within noise, +2.2 pts
Large option sets Choices with dozens of options to pick from.	1,677	0.806 0.787 to 0.825	0.775 0.755 to 0.796	0.796 0.776 to 0.816	Within noise, +3.1 pts
Long inputs Long threads and documents with the answer deep inside.	1,147	0.835 0.810 to 0.858	0.853 0.832 to 0.874	0.827 0.801 to 0.851	Within noise, −1.8 pts
Multilingual The same decisions asked in many languages.	1,812	0.796 0.776 to 0.814	0.782 0.762 to 0.802	0.775 0.755 to 0.793	Within noise, +1.4 pts
Rules and policies Apply a written policy to a case and decide.	1,759	0.891 0.875 to 0.907	0.831 0.812 to 0.849	0.872 0.855 to 0.890	MoE leads, +6.0 pts
Prompt injection Inputs that try to talk the model into the wrong answer.	798	0.764 0.734 to 0.793	0.600 0.565 to 0.634	0.747 0.718 to 0.777	MoE leads, +16.4 pts
Photos Choice and yes/no questions about real photos.	1,668	0.918 0.906 to 0.930	not supported	0.900 0.887 to 0.914	Image suite, Jev reads text only
Rendered scenes How likely a statement is, judged from a computer-rendered scene.	450	0.687 0.643 to 0.732	not supported	0.620 0.575 to 0.665	Image suite, Jev reads text only

Preliminary results, October 2026. 15 text suites, 25,134 items, development splits. 95% intervals for the DecisionNode models and Jev come from a bootstrap that resamples whole inputs. Jev (jev-1.13.0) was called through its official API; open models ran as their authors serve them. DecisionNode models are release candidate 1.

“Leads” and “trails” are used only where the two 95% intervals do not overlap. Source: decisionnode.com/benchmarks. DecisionNode is not affiliated with TypeSafe AI, Cloudflare or the other model publishers. Their names identify the systems measured.

Price

Per input token, DecisionNode MoE is priced the same as Jev: $0.042 per million. DecisionNode Flash costs $0.035, about 17% less. Output is free on all three. DecisionNode adds a batch price for work that can wait a few hours ($0.0294 on MoE, $0.0175 on Flash), a price for images and video ($0.09 and $0.08 per million tokens) and a free daily tier.

Each model counts tokens with its own tokenizer, so the same request can count differently on each. Both APIs report input tokens with every answer; send a sample of your real requests to both and compare. The pricing page has worked examples.

The API, side by side

The two requests have the same outline: a state and a map of typed questions with instructions and criteria. DecisionNode documents its own API only, so moving a Jev integration means pointing it at /v1/decide, using a DecisionNode key and model name, and reading yes/no answers from truth. For every field of Jev's request and response, see how the Jev API works.

Part	Jev	DecisionNode
Host and path	`api.typesafe.ai/v1/systemone`	`api.decisionnode.com/v1/decide`
Yes/no type	`noul`	`truth`
Model names	`jev-latest, jev-1.13.0`	`decisionnode-1.0, decisionnode-1.0-flash`
Images and video	`None`	`images and videos fields beside state`
Other endpoints	`GET /v1/models`	`GET /v1/models, uploads, batches, sessions`

A support-triage decision on DecisionNode

curl https://api.decisionnode.com/v1/decide \
  -H "Authorization: Bearer $DECISIONNODE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "decisionnode-1.0",
  "state": "We were charged twice for the October invoice (INV-20931). The second charge pushed our card over its limit and our team seats were suspended this morning. Please refund the duplicate and restore access.",
  "questions": {
    "route": {
      "type": "choice",
      "instructions": "Which team should handle this ticket?",
      "criteria": {
        "billing": "Charges, invoices, refunds and plans",
        "bug": "Something in the product is broken",
        "account": "Login, seats and permissions",
        "feature_request": "Asks for something the product does not do"
      }
    },
    "urgency": {
      "type": "score",
      "instructions": "How urgent is this ticket?",
      "criteria": [
        "routine: no deadline",
        "soon: within a few days",
        "today: blocks work today",
        "critical: outage or data loss"
      ]
    },
    "refund": {
      "type": "truth",
      "instructions": "Is the customer asking for money back?",
      "criteria": {
        "true": "Asks for a refund or a credit",
        "false": "Any other request"
      }
    },
    "escalate": {
      "type": "truth",
      "instructions": "Should a person handle this now instead of an automated reply?"
    }
  }
}'

This request and four more are on the use cases page.

Who should choose which

Choose Jev if

Your inputs are English text or JSON and need only choices, scores and yes/no.
Answers depend on general knowledge, where Jev scores 0.820 against 0.698.
You want an official Python or JavaScript SDK with retries built in.
Your customers' data must never be used for training, with no setting to change.

Choose DecisionNode if

You decide on photos, screenshots, documents as images, or video frames.
You need a number, a point or a box as the answer.
The state comes from users, web pages or agents that may try to steer the answer.
You run large jobs that can wait hours, at 30% to 50% off.
You want to start free: 50 DecisionNode Flash requests, 20 DecisionNode MoE requests and 10 booster calls a day.

Jev vs DecisionNode questions

Is DecisionNode more accurate than Jev?

On the mean of 15 text suites, DecisionNode MoE scores 0.816 and Jev 1.13 scores 0.803 (preliminary, October 2026). MoE leads clearly on prompt injection, web tasks and rules and policies; Jev leads clearly on knowledge questions and new task families; the other ten suites are within noise.

Does DecisionNode train on my requests?

By default, yes: DecisionNode uses requests for training unless you opt out, and zero data retention is available on every plan. TypeSafe states that Jev is not trained on customer requests, and offers zero retention to enterprise customers.

Is DecisionNode faster than Jev?

No head-to-head speed figure is published. DecisionNode publishes one speed result: Flash answers a short decision in 19 ms end to end (about 5 ms on the server), against 1,300 ms for GPT-5.6 Luna timed from the same client.

Does DecisionNode read images?

Yes. A request can carry up to 16 images and up to 4 videos as frames, and DecisionNode MoE scores 0.918 on the Photos suite. Jev reads text only.

Run your own decisions on DecisionNode

Send text, JSON or images with typed questions and get probabilities back. Free tier: 50 Flash and 20 MoE requests a day.

Try DecisionNode