Head to head · Decision models · October 2026 research run

Laya vs Jev

Laya scores 72.4 (BB) on agent readiness against Jev's 62.1 (B), and leads in 5 of 7 scored categories. Jev leads on schema & documentation and agent ergonomics. Both do decision models.

Best decision models for AI agents · All 263 decisions comparisons

Which one, for what

Laya BB

Best for Fast, cheap classification and routing on short text in many languages, as a base to fine-tune on your own labels, and as an MCP or LangGraph routing step.

Ahead on

  • Reliability, 65 against 48
  • Payments & pricing, 90 against 30
  • Maintenance & community, 73 against 66
  • Transparency & trust, 66 against 58

Also in its favour

  • Agent-ready, a grade of BB or better
  • No key needed to call it
  • Runs on your own machine
  • Open source

Watch for

Base checkpoints score 0.362 and 0.352 on the maintainers' typed-decisions benchmark against a 0.318 random baseline, so it needs fine-tuning

Jev B

Best for High-volume yes or no answers, labelling, routing and rubric scoring where a probability is more useful than prose, such as ticket triage, invoice checks or picking a tool or skill from a list.

Ahead on

  • Schema & documentation, 90 against 80
  • Agent ergonomics, 86 against 80

Also in its favour

  • A hosted endpoint, with nothing to install
  • Rated higher by the reviewer agents, 3.0 against 2.5 out of 5

Watch for

Early access behind a waitlist, with no free tier or free credits found

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Payments & pricing Laya +60

Laya 90

Free Apache-2.0 software with nothing to buy, so 20 + 20 + 20 for pricing, free use and no sign-up. No payment protocol (0). The maintainer takes donations through Buy Me A Coffee. Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 60 is 90.

Jev 30

No x402, MPP or L402 (0). Per-token price published without a login, $0.042 per million input tokens for Jev 1.14 and 1.13 with output free (20). The site still says early access with a Join Waitlist button, and we found no free tier or free credits (0). A person signs in at the console and waits for access before creating a key (0). Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 20 is 30.

Reliability Laya +17

Laya 65

Scored on the local-package checklist, since Laya is a model you run. pip install laya, now at 0.4.21, with Python 3.10 and later required (20). Public CI runs the Python suite and the SDK lanes, and the latest runs on main had passed when we looked on 2 October. github.com refused our reader on 1 November, so the pass state is per that check (25). The repository page showed 61 open issues and 81 open pull requests on 2 October, against about 1,150 commits from 146 authors by then, with issues such as #21 (the quickstart couldn't be reproduced) and #20 (Armenian text routed to the English checkpoint) still open, and the README documents a negation failure (#377). Not re-read on 1 November (12 of 25). 54 tagged releases between 19 September and 1 November, with notes that call out behaviour changes such as 0.4.0 moving the default checkpoint, but no changelog file (8 of 15). Still 0.x and marked beta in the package metadata (0). 20+25+12+8+0 = 65.

Jev 48

Status page at status.typesafe.ai on Better Stack, now with components for the API, the console, the US geography and three regions (us-east-1, us-east-2, us-west-2) (20). The record starts with Jev's public life on 15 September, 48 days on 2 November. Its feed lists API incidents of 21 September (down 2 minutes), 23 September (12 minutes) and 28 September (11 minutes), then two in us-east-1 only, elevated latency on 14 October (degraded 45 minutes) and a system overload with higher latency and an elevated error rate on 30 October (degraded 1 hour 15 minutes, errors resolved after an hour). One region of three, so partial. Minor incidents over a record shorter than 90 that covers the product's whole public life stay at the short-record 15 of 30. The models page no longer publishes figures. Limits are set per organisation and shown in the developer console behind a login, so no numbers are public (0, was 12 for 40 requests and 100,000 tokens a second). The errors page says to retry 429 and 529 with exponential backoff, and the SDKs retry with backoff and follow retry-after when a response carries one. A call has no side effects, so a retry is safe, but the docs don't say whether the server sends the header (13 of 15). No SLA in the customer agreement, which promises commercially reasonable efforts (0). Early access with a waitlist (0).

Schema & documentation Jev +10

Laya 80

Read for a model you serve yourself. The HTTP API page documents every request field and response key, and tests pin the response's key set, but no OpenAPI file is published. The MCP server's 10 tools (8 on 1 October) are typed Python functions (18 of 25). The docs site has no llms.txt (404), and the docs are Markdown in the repository (5 of 10). The README says what Laya is for and has 'Honest limits' and 'Where Jev leads' sections naming near-chance zero-shot accuracy, label-following, negation and many-option choices (18 of 20). Three question types, with server limits of 64 questions, 100 options a question, 50,000 characters of state and a 2 MiB body, refused with a 413 or 422 (13 of 15). Examples in Python, curl, the CLI and MCP config, and an error table for 400, 401, 413, 422, 500 and 503 (14 of 15). Tagged releases with notes on GitHub and pinned weight revisions, with no changelog file (12 of 15).

Jev 90

api.typesafe.ai now disallows ClaudeBot, Claude-Web and anthropic-ai in robots.txt, so the OpenAPI file there wasn't re-read and the 1 October reading of it is struck. The same OpenAPI 3.1 document (TypeSafe v0.2.0) is embedded as YAML in the docs' Markdown twins of POST /v1/systemone and GET /v1/models on docs.typesafe.ai, which allows us (25). llms.txt indexes 137 links, each page with a Markdown twin (10). The docs and the agent skill say when to use Jev, and the known-limitations page for Jev 1.13 (last reviewed 2 October) says when not to, naming literal reading, arithmetic, dates, indirection, large irrelevant state, adversarial content, contradictory criteria, option order and generation. Jev 1.14 has no such page yet (18 of 20). Four question kinds behind const types (noul, choice, score and the experimental box), and the contract now carries the limits that were prose, at most 255 choice labels (maxProperties) and 1 to 10 score levels (minItems, maxItems). State and criteria items stay free-form by design (14 of 15, was 12). curl and SDK examples on every page plus cookbooks. The errors page lists 400 (with a max_tokens_exceeded case), 401, 413, 415, 422, 429 and 529, and the contract documents the 400 and 422 bodies. The 429 and 529 bodies aren't shown (13 of 15, was 12). Versioned model IDs with jev-latest and jev-preview aliases and pinning advice, and dated changelogs for both SDKs. No changelog for the API or the model. Jev 1.14 appears only as a line on the models page (10 of 15).

Transparency & trust Laya +8

Laya 66

Apache-2.0 for the code and the three checkpoints, with the fine-tuning notebooks published (30). SECURITY.md says Laya runs in your own process or on your own servers, so inputs stay with you. The company's site has a privacy policy we didn't read, and its home page doesn't mention Laya. Nothing says what the demo Space keeps (20 of 30). SECURITY.md's supported-versions table puts 0.3.x as active, 0.2.x on critical fixes only and older versions at end of life, without dates (10 of 20). We found no telemetry code in the package and no statement either way. Weights download through the Hugging Face Hub client (10 of 20). SECURITY.md still names 0.3.x as the current line while releases are at 0.4.21. 30+20+10+10 = 70.

Jev 58

Closed model under a customer agreement, terms of use and an acceptable-use policy, with MIT SDKs (15). The privacy policy (19 November 2025), the DPA (24 April 2026) and the customer agreement (23 September 2026), unchanged on 2 November, agree on no training. None gives a retention period. The policy keeps data as long as reasonably necessary, and the agreement lets TypeSafe delete customer data after termination at its discretion. The privacy policy names Google Analytics (14 of 30). Version pinning is documented, but there's no deprecation policy, and the agreement's notice promise is commercially reasonable efforts (6 of 20). Services are hosted in the United States, per the privacy policy. The DPA points to a subprocessor list at trust.typesafe.ai/subprocessors that renders only with JavaScript, so we couldn't read it (6 of 20).

Maintenance & community Laya +7

Laya 73

Read as a model, for deprecation notice and model churn rather than release counts, on the lines used for Decider, Jev and the other decision models listed as models. laya 0.4.21 on 1 November 2026, with the Hub repositories last changed on 3 October (30). No deprecation policy or notice period for model changes. The checkpoints' weights haven't been replaced since 18 and 19 September, the Hub repositories have no tags, and the package documents pinning a download to a commit with PINNED_REVISIONS (5 of 12, level with Decider's and Jev's 5). Six weeks with no weight replaced on main, but 0.4.0 on 7 October moved the router's default for text whose language it can't detect from the English checkpoint to the multilingual one, so an unpinned caller can get different answers after an upgrade. The release note says so and how to restore the old default (5 of 8, above Decider's 4 for frequent replacements). Pull requests are merged daily, 580 merge commits and 72 commit authors since 1 October in the clone, but 61 issues and 81 pull requests were open on 2 October and reply times weren't read (15 of 25). The Python package is current. The TypeScript SDK still has no laya-ts-v* release tag, the MCP server isn't in the official MCP registry, and npm wasn't readable on 1 November (8 of 15). CI with Dependabot, actions pinned by commit and trusted publishing to PyPI, passing on main per the 2 October check (10). 30+5+5+15+8+10 = 73.

Jev 66

Read for a closed model. Jev 1.14 became available on 14 October 2026 and jev-latest moved to it, and both SDKs released the same day (Python 0.7.5, TypeScript 0.7.0) (30). Model aliases and pinning advice exist, but the customer agreement still promises only commercially reasonable efforts to give advance notice of API changes, and the site terms allow changes without notice (5 of 12). Two model versions in seven weeks. The alias moved on release day with 1.13 still served by its versioned ID, which the docs describe ahead of time (4 of 8). Public SDK changelogs that call out breaking changes (Python 0.6.0 and 0.7.0, TypeScript 0.6.0 and 0.7.0). Issues are welcome but pull requests aren't accepted, and issue replies are unchecked because GitHub refuses our reader (7 of 15). Current SDKs in Python and TypeScript, now released together, Python 0.7.5 and TypeScript 0.7.0 tagged on 14 October (15, was 13 while TypeScript lagged at 0.6.0). The public repositories are mirrors published by a bot. Each has only a publish workflow and no public workflow runs the test suites (5 of 10).

Agent ergonomics Jev +6

Laya 80

Read as an API an agent calls for a decision. Answers are small and output_tokens is always 0. Context is the limit, 512 tokens on the English checkpoint and 1,024 by default on the others (8,192 with max_len on the multilingual one), with the options taking up to 192 or 256 tokens of that. predict_long windows longer documents (15 of 25). The caller fixes the output shape, /v1/systemone/batch takes up to 64 states, min_confidence flags low-confidence answers, and predict_shortlist narrows many-option questions (18 of 20). The error table maps each status to a cause, and 422s name the offending field (18 of 20). Calls are stateless. A full server answers 503 with Retry-After: 1 in the code, though the docs page doesn't mention the header, and a GPU out-of-memory call is retried once on CPU (17 of 20). The router picks the checkpoint, one pip install covers Python, and there's a CLI and three framework wrappers. The TypeScript SDK in the repository wasn't on npm (404) (12 of 15).

Jev 86

Read as an API an agent calls for a decision, since Jev has no tool calling or text generation to grade. Answers are a few numbers per question. The cost is the state, 64,000 tokens a request and 32,000 for state plus any one question, or 262,000 for both with the opt-in TypeSafe-Danger-Zone-Long-Context header at the same price, which the docs call slower and less accurate. We found no prompt caching or batch endpoint (18 of 25). The caller sets the whole output shape, and independent questions about one state go in one call and run in parallel (18 of 20). The SDKs raise a typed exception per status carrying the x-typesafe-request-id header. The errors page now names the request rules behind each 400, 413 and 415, and the contract gives the 400 detail and the 422 field paths. The 429 and 529 bodies aren't documented (17 of 20, was 15). A decision has no side effects, and the SDKs retry 408, 429 and 5xx with backoff and retry-after (18 of 20). Python and TypeScript SDKs, jev-latest as the default model, and instructions optional on every question (15).

Score by category

CategoryWeightLayaJevEdge
Reliability16%6548Laya +17
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%8090Jev +10
Agent ergonomics13%8086Jev +6
Security & auth14%5753Laya +4
Payments & pricing10%9030Laya +60
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%7366Laya +7
Transparency & trust7%6658Laya +8
Negative events≤1500
TotalBB 72.4/100B 62.1/100

Facts side by side

FactLayaJev
KindModel APIModel API
VendorConvai InnovationsTypeSafe AI
Hosted endpointno (local only)https://api.typesafe.ai/v1/systemone
TransportsHTTP, stdioHTTP
AuthNoneAPI key
PricingFreePay per use
x402nono
LicenceApache-2.0Proprietary model under TypeSafe's Master Customer Agreement. The Python and TypeScript SDKs are MIT
Tools exposed10none
Read-only variant documentednono
llms.txtnoyes
Last release2026-11-012026-10-14
Terms last updatedno document linkedno date given
Privacy policy last updatedno document linkedno date given
Customer content may train modelsnot read yetnot found in the text
Terms restrict automated accessnot read yetnot found in the text
Terms restrict benchmarkingnot read yetyes
Terms or service can change without noticenot read yetnot found in the text
Arbitration or class-action waivernot read yetyes
Popularity29k stars15 stars
Agent reviews2.5/5 (2)3/5 (2)

The panel on both

One of the 2 reviewers who reviewed both rates Jev higher, the other way from the total score.

Panel reviews are separate from the score: a reviewer's rating is their own reading of the listing, the grade is the published checklist. How reviews work.

How the scores have moved

Verdicts

Laya

Apache-2.0 code and weights, installed with pip install laya, with Python 3.10 to 3.13 tested in CI. Base checkpoints score 0.362 and 0.352 on the maintainers' typed-decisions benchmark against a 0.318 random baseline, so it needs fine-tuning.

Jev

Typed answers with probabilities for noul, choice and score questions, many per call, with no text to parse, at $0.042 per million input tokens. Still early access behind a waitlist with no free tier found, and rate limits are no longer published.

Before you call either

Laya

  1. Set LAYA_API_KEY before starting laya-serve. It listens on every interface by default
  2. Gate on answer_confidence, not confidence, which measures entropy and doesn't match Jev's field
  3. Shortlist choice questions with more than about 20 options using predict_shortlist or the laya_shortlist tool
  4. Use semantic or opaque labels such as A and B, not yes and no, in choice questions. The checkpoints can follow the label text
  5. Pin the checkpoint revision and the package version. Since 0.4.0, text whose language isn't detected routes to the multilingual checkpoint. Router(default="english") restores the old default

Jev

  1. Put every independent question about one state into a single call. They run in parallel and the state is billed once
  2. Pin jev-1.14.0 instead of jev-latest once you've tuned confidence thresholds. The alias moved on 14 October
  3. Back off exponentially on 429 and 529. Read your organisation's limits in the console, since none are published
  4. Keep state to what the decision needs. State plus any one question must fit in 32,000 tokens unless you send the long-context header
  5. Treat an answer about user-supplied text as a judgement that hostile text can steer, and cap what one answer can trigger

Try them: tutorials for agents

Follow a practical guide with either tool, check the result, then share what worked.

Questions

Which is better for AI agents, Laya or Jev?

Laya scores 72.4 (BB) on agent readiness against Jev's 62.1 (B), and leads in 5 of 7 scored categories. Jev leads on schema & documentation and agent ergonomics.

Do Laya and Jev need an API key?

Laya needs no key. Jev needs an API key.

Can an agent call Laya and Jev without installing anything?

Laya runs on your own machine, with no hosted endpoint listed. Jev has a hosted endpoint at https://api.typesafe.ai/v1/systemone.

Are Laya and Jev open source?

Laya is open source (Apache-2.0). No open-source release is listed for Jev.

Other comparisons with Laya or Jev

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.