Jev vs DOPEST 1.0 31B: published benchmark comparison

Data: JevBench v1.7.4, published 2026-10-25 · method v1.7 (10 s rule), re-scored from the recorded per-decision times

Rank #3 on the open-weights board in v1.7.4.

Compare Jev-compatible models by separate measures and their measurement conditions. A historical measurement is not a current rank.

Jev and DOPEST 1.0 31B: published axes and category radars

Pick any two. Radars for the score axes, capability by subject topic and use cases, chance-corrected competence per request type on the open and sealed sets, and competence per tier on each set. Further out is better on every spoke; the link keeps the pair.

System A

System B

A: DOPEST 1.0 31B — Open weights · LLM decoder · Score 67.8 (#3)
B: Jev 1.13.0REFERENCE · TypeSafe — Jev — reference (TypeSafe, closed) · Score 71.4 (ranked, not ranked)

The four score axes

0–100, the values in the table. An axis a system has no published value for is left as a gap (it counts as 0 in the composite).

Capability by subject topic

DOPEST 1.0 31B: No per-category values are published for this system.

Linear signed scale: −100 at the centre, 0 on the bold middle ring, 100 at the rim.

What the decision is about — math and dates, coding, rules, money, support work, everyday messages, safety. Chance-corrected competence per category (0 = at or below chance; negative averages are reported as 0, 100 = perfect), DOPEST 1.0 31B: open+sealed; Jev 1.13.0: S+P+L1+L2+L3 items pooled; only categories with at least 30 items are spokes, smaller ones are listed below. Hover a category for its definition and item count.

What each category means · items per category

Rules, policy & law — applying written rules: company policies, contracts, regulations, eligibility. 1616 items (151 open / 1465 sealed)
Coding & software — code, SQL, repositories, developer tools and IT systems. 382 items (62 open / 320 sealed)
Math & numbers — a calculation decides the answer: arithmetic, word problems, probability, dates, units. 370 items (37 open / 333 sealed)
Finance & commerce — money: payments, refunds, invoices, orders, expenses, insurance payouts. 223 items (12 open / 211 sealed)
Support & operations — support tickets, incidents, logistics, scheduling desks and routing work to a team. 160 items (21 open / 139 sealed)
Everyday language — short everyday messages: intents, assistant requests, reading a detail out of a text. 145 items (11 open / 134 sealed)
Safety & security — untrusted or injected instructions, fraud, moderation, access and security triage. 109 items (6 open / 103 sealed)

Use cases (TypeSafe categories)

DOPEST 1.0 31B: No per-category values are published for this system.

Linear signed scale: −100 at the centre, 0 on the bold middle ring, 100 at the rim.

The real-world application area of each decision, after the TypeSafe use-case map; an item can count in two. Chance-corrected competence per category (0 = at or below chance; negative averages are reported as 0, 100 = perfect), DOPEST 1.0 31B: open+sealed; Jev 1.13.0: S+P+L1+L2+L3 items pooled; only categories with at least 30 items are spokes, smaller ones are listed below. Hover a category for its definition and item count.A point on the bold 0 ring printed 0.0 is a measured value, not a gap: these published cells clip below-chance results to 0, so 0 means at or below chance. Their score markers cannot fall inside that ring.

What each category means · items per category

Model routing — deciding which LLM or model tier should handle a prompt; classifying a request's intent, domain, difficulty or risk to pick a model. 524 items (86 open / 438 sealed)
Legal and compliance — contracts, policies, regulations, eligibility rules, checking documents against written requirements. 515 items (74 open / 441 sealed)
Customer support — classifying tickets and customer intents, urgency, refunds, routing cases to a team, checking support replies. 240 items (37 open / 203 sealed)
Other — none of the above: general reasoning, arithmetic or date puzzles, everyday assistant requests without a business workflow. 174 items (24 open / 150 sealed)
E-commerce marketplaces — product listings, product attributes, orders and deliveries, seller catalog policy. 119 items (9 open / 110 sealed)
Insurance claims — first-notice-of-loss reports, claim complexity, missing information, claim triage. 119 items (9 open / 110 sealed)
Risk assessment — estimating probabilities and severity of risks, incidents, vendor or underwriting risk. 118 items (8 open / 110 sealed)
Financial crime — suspicious transactions, KYC, fraud alerts, entity matching for investigations. 101 items (9 open / 92 sealed)
Feature extraction for predictive modeling — turning natural-language data into probabilistic features or estimates for a downstream prediction. 99 items (5 open / 94 sealed)
Lead generation — matching companies or inbound messages to an ideal customer profile, buyer intent, prioritising and routing sales leads. 95 items (4 open / 91 sealed)
Recruiting — resumes, applications, interview feedback, matching candidates to roles. 94 items (1 open / 93 sealed)
Graphs and knowledge graphs — entity types, relationships between records, contradictions between facts, multi-step lookups across linked facts. 94 items (3 open / 91 sealed)
LLM guardrails — checking LLM inputs, outputs and tool calls: prompt injection, jailbreaks, policy violations, tool-call errors, answer quality. 93 items (5 open / 88 sealed)
Moderation and trust and safety — user content moderation: toxicity, harassment, spam, unsafe advice, personal data exposure. 93 items (7 open / 86 sealed)
Semantic code linting — checking code or writing against conventions and guidelines, as in CI review of code. 90 items (3 open / 87 sealed)
Search and retrieval — scoring query-to-candidate relevance, reranking results, selecting useful context for RAG. 90 items (5 open / 85 sealed)
Scientific discovery — screening papers, labelling research passages or survey answers, checking that citations support claims, methodology checks. 90 items (2 open / 88 sealed)
Advertising — ad creatives, campaign copy, brand safety, prohibited claims in ads. 88 items (3 open / 85 sealed)
Gaming — player reports, in-game chat, game support. 86 items (3 open / 83 sealed)
Demand forecasting — purchase intent, product interest, demand and supply signals for forecasting. 83 items (3 open / 80 sealed)

Competence per request type, open / sealed

Linear signed scale: −100 at the centre, 0 on the bold middle ring, 100 at the rim.

Chance-corrected competence for Choice, Noul and Score on the 300 open and 1200 sealed decisions.

Competence per tier — open set

Linear signed scale: −100 at the centre, 0 on the bold middle ring, 100 at the rim.

Per-tier competence, the three request types pooled by their published decision counts.

Competence per tier — sealed set

Linear signed scale: −100 at the centre, 0 on the bold middle ring, 100 at the rim.

Per-tier competence on the sealed decisions, types pooled the same way; item text stays private.

How the categories were made

O1S chance-corrected competence: equal mean across the request types present, clipped to 0–100. Zero means the equal-weight mean across present request types was at or below baseline; negative means are clipped. It does not mean every type scored zero. Choice uses a random-option baseline; Noul uses a fixed 50% accuracy baseline (abstentions count as wrong); Score uses the midpoint-guess error (random-guess fallback when every gold is at the midpoint). Compare models within a category; row pools and supported types can differ.

Categories use ruled subject-topic labels and authoring use-case precedence across the answered pool union, with stable item identities counted once.

Official O1S; recorded failures count where supported; no missing response is imputed; stable identity dedupe parent first.
S-based category cells pool recorded supported responses from S+P and the answered L1/L2/L3 supplements. API overlay category cells keep their separately recorded A4/A5+P union. Headline and per-type/tier scores are unchanged.

All values as a table

Spoke	A: DOPEST 1.0 31B	B: Jev 1.13.0
The four score axes
Intelligence	73.9	63.6
Calibration	90.2	90.6
Speed	82.0	91.2
Cost	49.2	54.7
Capability by subject topic
Rules, policy & law	—	49.4
Coding & software	—	64.3
Math & numbers	—	10.4
Finance & commerce	—	40.9
Support & operations	—	48.8
Everyday language	—	77.5
Safety & security	—	52.2
Use cases (TypeSafe categories)
Model routing	—	73.0
Legal & compliance	—	68.6
Customer support	—	49.6
Other	—	40.1
E-commerce	—	40.5
Insurance claims	—	46.1
Risk assessment	—	32.8
Financial crime	—	35.8
Feature extraction	—	28.3
Lead generation	—	49.2
Recruiting	—	0.0
Knowledge graphs	—	28.8
LLM guardrails	—	68.0
Moderation	—	51.6
Code linting	—	29.1
Search & retrieval	—	80.9
Science	—	21.2
Advertising	—	16.4
Gaming	—	5.1
Demand forecasting	—	0.0
Competence per request type, open / sealed
Choice · open	80.8	79.0
Choice · sealed	79.3	75.8
Noul · open	81.8	54.5
Noul · sealed	77.8	46.1
Score · open	67.0	63.3
Score · sealed	56.5	63.0
Competence per tier — open set
Easy	95.4	93.9
Standard	72.5	63.2
Judge	72.5	67.7
Hard	79.5	65.6
Competence per tier — sealed set
Easy	88.0	78.4
Standard	73.9	68.7
Judge	71.2	66.9
Hard	70.7	58.9

Published values

Measure	Jev 1.13.0 (TypeSafe AI)	DOPEST 1.0 31B
Capability Score	77.1	82.0
Rank (open-weights board)	Reference, not ranked	#3
Composite (secondary)	71.4	67.8
intelligence	63.6	73.9
calibration	90.6	90.2
speed	91.2	82.0
cost	54.7	49.2
USD per 1,000 decisions	$0.032 per 1,000 decisions	$0.049 per 1,000 decisions
p50 latency (seconds)	0.2	0.236 measured on-GPU, no network · 0.62 est. incl. network (2x + 150 ms)
Method	v1.7 · 10 s rule	v1.7 · 10 s rule
p95 (measured) · decisions over 10 s	p95 0.311 s (measured) · 0 of 1,500 over 10 s	p95 0.435 s (measured) · 0 of 1,500 over 10 s
Before the 10 s rule (v1.6)	v1.6.43: A 71.49 · #18	—
Last measured	2026-10-05	2026-10-25

* Self-hosted rows: the first value is the p50 we measured on our own hardware over loopback (on-GPU, or on-CPU where marked, no network). It is not directly comparable with API latency, which we time end to end from our measuring host, network included. “est.” is the est. incl. network (2x + 150 ms) that the ranking and the latency cap use for every self-hosted row. Model authors can ask us to time a hosted endpoint instead (on request).

Jev 1.13.0 (TypeSafe AI)

proprietary API

Published source

Measured on: the operator's hosted API
Cost basis (not published): TypeSafe list tariff USD 0.042 per 1M input tokens, output free; measured on the v1.6.1 common cost basis (all items except those outside the Jev reference input range) from this run's token usage
Latency adjustment: none (API)

DOPEST 1.0 31B

Apache-2.0 adapter per the model card; base google/gemma-4-31B-it public, ungated; author serving code public at the pin until 25 Oct 2026, repository private since then (no serving code redistributed here)

Published source

Measured on: Evaluator-owned rented Verda GPU VM, one NVIDIA A100-SXM4-80GB (81,920 MiB, driver 580.178.04), FIN-03 (Finland, EU), PAYG non-spot, USD 2.2311917808219177/h recorded by the runner, 300 GB OS volume. One VM for this row's measured attempt. Measured 25 Oct 2026, VM window about 17:22-18:03Z, about USD 1.53 GPU rental (elapsed time at the recorded hourly rate; attempts 1 and 2 are not included in that figure). Author server at 127.0.0.1:8000 in a --network none --read-only, cap-drop-ALL container started through env -i: read-only code, model and input mounts, tmpfs caches, private output, empty provider credentials (DEEPSEEK, OPENAI, HF, MODAL, OP) and HF offline flags. Before sealed input: in-container network cut proven (only loopback, no IPv4/IPv6 default route, DNS and TCP fail); model files re-hashed (12 files, 0 failures); a synthetic pre-warm passed on the first try; the TTL fit gate admitted the run. Input sha256 dfd36720077ef7c41ace367524896507c7f44ce71306fb3e659e8b85069d995b was re-checked inside the container (match). Complete 1,500-row run, rc 0; scored loop 516.0 s. Raw output sha256 33fe1dc0d85a195fe0bd65d1d0a0fb21c9d78abc1d6c72631eceec306bf33ee8 (the pull-hash receipt). Node-shred and local-shred receipts verified; exact-id teardown receipt confirms the VM closed. GPU rental is separate from the benchmark's estimated model tariff.
Cost basis (estimated): ESTIMATE (*est.), base-model market reference through the frozen pricing_v15 module (sha256 aa2afcd5a58d0c4cc0e3d6867d590c70b6ac58e971410c27faf9628f97bcf8b7, snapshot of 22 Oct 2026): base string google/gemma-4-31B-it (the adapter config's base_model_name_or_path, in the author's capitalised spelling) priced as google/gemma-4-31b-it USD 0.09/M input and USD 0.34/M output; the output term is zero because output usage is zero. The exact case of 31B is required by the frozen module: it resolves to the lowercase google/gemma-4-31b-it listing, while the all-lowercase spelling raises 'base model lacks an explicit market-listing decision' here. No explicit reference or override is used. Estimated, not charged, not a quote from the author or any host. Tokens as reported natively by the usage block of the author's server: 990,293 input and 0 output tokens across all 1,500 answered items; no generated text. Native usage equals our first-party offline count exactly (0 mismatches over 1,500 items; our sealed-input count was also 990,293 and the prompt-level concordance had 0 mismatches). The 15 items that the author's 12,288-token prompt limit truncated are answered and are INCLUDED in the usage and in the cost (their counted input is the truncated prompt, at most 11,985 tokens). Under the zero-output convention for native-probability rows (the published native-probability convention, as applied to Vera and Turion) the output term contributes nothing. Measured arithmetic: USD 0.059417580 per 1,000 decisions over all 1,500 rows = 0.919852886x the unchanged USD 0.06459465517241379 cap; separately USD 0.049372545 per 1,000 on the common 1,485-item basis (814,647 input tokens; the same items for every token-priced row, which excludes the draw's 15 standard long positions) = 0.764344129x. WITHIN COST CAP on both bases: within the automatic cost treatment; no cap exception or top-up. The all-1,500 figure is close to the cap (0.92x) while the common-basis figure is 0.76x, so the measured number, not the pre-run forecast, decides.
Latency adjustment: x2 + 0.15 s (assumption, not measured)

Frequently asked questions

What does JevBench show for Jev and DOPEST 1.0 31B?: Jev 1.13.0 has Capability 77.1 and is the unranked reference (the open-weights board ranks 162 open-weights systems; hosted API offerings are ranked on the API leaderboard). DOPEST 1.0 31B has Capability 82.0. Rank #3 on the open-weights board in v1.7.4. Historical scores use their original method and are not ranked against the current release.
Is there a time limit per decision?: Yes, from JevBench v1.7.0 (method v1.7): a decision that takes longer than 10 seconds (measured time) counts as a wrong answer and as zero calibration for that item. Each row shows its p95 and how many decisions were over the limit; the v1.6 values stay visible.
How does Capability differ from Composite?: Capability is the mean of Intelligence and Calibration; headline ranks require both official cost and median-latency caps. Composite additionally scores Speed and Cost and is secondary.
Can I compare the prices as actual bills?: Jev 1.13.0 (TypeSafe AI): not published, $0.032 per 1,000 decisions. DOPEST 1.0 31B: estimated, $0.049 per 1,000 decisions. Estimates and self-reported vendor prices are not measured charges.
Is DOPEST 1.0 31B open source?: The published licence note for DOPEST 1.0 31B is Apache-2.0 adapter per the model card; base google/gemma-4-31B-it public, ungated; author serving code public at the pin until 25 Oct 2026, repository private since then (no serving code redistributed here). Jev 1.13.0 is listed as proprietary API. Check the linked sources for terms.