Skip to content
Made with Jev

Cloudflare Clef alternatives

Looking for alternatives to Cloudflare Clef? TypeSafe Jev is generally available with the same question types (Choice / Score / Noul), documented pricing, proof builds and, since 14 October 2026, image input in preview. When Clef still wins — on audio and video, open weights and, with Clef-flash, the lowest price per token — and where each fits.

Updated · by

In short

  • Clef: System One on Workers AI, multimodal (text/JSON/images/video), open weights, $0.24/M input. Clef-omni adds audio at $0.15/M; Clef-flash is $0.038/M with a 24k hosted context.
  • Jev: GA, text plus PNG/JPEG images in preview, $0.042/M input (output free), madewithjev proof builds, no open weights.
  • Both handle same question types (Choice/Score/Noul), ≤64 questions per call, context to 65k tokens.
  • Clef wins: audio and video, production vision, self-host weights, Cloudflare edge, and the cheapest token with Clef-flash. Jev wins: documented builds, TypeSafe GA path, more context than Clef-flash.

Who this page is for

You are here because Cloudflare Clef locks you into Workers AI billing, or you want TypeSafe GA without Cloudflare accounts, or you need smaller local open-source alternatives, or you are comparing per-token pricing across decision APIs.

This page covers the documented alternatives. Jev first (most directly comparable GA peer), then when Clef still fits, then Perplexity and OpenAI peers, then self-hosted open-source paths.

Primary alternative: TypeSafe Jev

TypeSafe Jev launched September 15, 2026 as a System One decision model. Same category as Clef: Choice, Score and Noul question types, batch up to 60+ questions in one call, confidence scores and per-option probabilities.

What Jev does that Clef docs do not mention

  • Lower per-token cost than Clef and Clef-omni: $0.042 per million input tokens (output free) vs $0.24/M for Clef and $0.15/M for Clef-omni. Clef-flash is cheaper at $0.038/M since Cloudflare's 9 October price change, but its hosted context window is 24k against Jev's 64k per request.
  • A Box answer: Jev 1.14 returns one bounding box on an image, in preview. Clef's docs do not list one.
  • Proof builds: madewithjev catalogs real working builds with published cost/latency figures, not vendor claims.
  • GA with no Cloudflare account: sign up at console.typesafe.ai, no Workers billing.

Sources: TypeSafe models, TypeSafe image docs

Builds shipping on Jev today

Duncan on X

Duncan

@ephraimduncan

Built a model router with Jev by @typesafeai. Jev decides what model fits your request best and the request is sent to that model.

Video from @ephraimduncan

XRouting and model choice

A model router on Jev

tamara on X

tamara

@tamarajtran

found the perfect use case for @typesafeai Jev: instant compaction in 2026, why is compaction still a summarization prompt? Jev can make it instant by scoring every tool call and dropping what’s irrelevant

Video from @tamarajtran

XContext and memory

Pick

Instant compaction for Claude

Gregor Zunic on X

Gregor Zunic

@gregpr07

Breaking: Browser Use + Jev = Ultrafast ⚡ Findings flights took 7s and cost only $0.0039 🤯 > new action space every step > DOM state space > small LLM fallback to type (this video is at 1x speed btw) Built a tiny open source browser agent. try it below ↓

Video from @gregpr07

XSearch

Pick

Flight search with Browser Use

Run time
~7 s
Cost
~$0.004
Hassan on X

Hassan

@nutlope

I used Jev to classify 1,018 AI research papers. The result: $0.08 total cost and 256ms median end-to-end latency per paper. The pipeline was: 1. Summarize each paper with DeepSeek V4 Flash 2. Send the title + summary + 24 possible topics to Jev 3. Use Jev to classify each paper 4. Visualize everything on http://1kpapers.com The summaries cost $3.99 on @togethercompute. The classifications cost $0.08 on @typesafeai. So for just over $4 of inference, I ended up with a pretty useful way to explore the top AI research papers from the past year. I think this is where things are heading: different models for different parts of the workflow, instead of using one model for everything. I’m running evals on the Jev classifications before replacing the current ones, but the site is already live: http://1kpapers.com

Video from @nutlope

Live siteDocuments and OCR

1kpapers

Documents
1,018 papers
Total cost
$0.08
Median latency
256 ms

More on the directory, grouped by job on Jev use cases.

When Clef is still the better fit

Audio, video and production vision. Clef accepts text, JSON, images and video per Cloudflare Workers AI docs, and Clef-omni adds audio. Jev reads PNG and JPEG images in preview only, behind a header, and takes no audio or video. If your workflow needs to classify video frames or sound, or images in production without preprocessing, Clef handles it.

Lowest price per token. Clef-flash costs $0.038 per million input tokens since 9 October 2026, below Jev's $0.042. The catch is context: the hosted Clef-flash window is 24k.

Open weights. Clef publishes Apache 2.0 weights on Hugging Face. TypeSafe has not released Jev weights. For self-hosted alternatives, Clef is the documented open path in this exact category; or see open-source Jev alternatives (Laya, Julia 1, kev, adapters).

Cloudflare edge. If you are already on Workers AI or want edge deployment, Clef is the native decision model. No need to wire an external API.

Other decision model peers

Clef is part of a category that emerged in September 2026. Brief notes on peers where docs exist; verify current state before choosing.

Perplexity Decisions API

GA at docs.perplexity.ai. Model pplx-decider-v1-27b, 1–128 questions, images via base64, $0.04 per million input tokens (output free). Slightly lower per-token than Jev; multimodal like Clef. See Perplexity Decisions alternatives.

OpenAI Decisions API

Announced at DevDay 2026, public beta on GPT-6 Luna, supports text or images. Schema and Decisions-specific pricing unpublished as of early October 2026. See OpenAI Decisions alternatives.

Open-source path

For local or self-hosted alternatives without vendor billing, see open-source Jev alternatives. Examples: Laya at 421M parameters (86.5 decisions/second, ~9 ms P50 per author), Julia 1 at 144M (runs on CPU), kev, and adapters that constrain next-token logits on models you already serve.

How to decide

  1. Do you need vision, audio or video? Clef or Perplexity document vision, and Clef-omni adds audio and video. Jev reads PNG and JPEG images in preview, with no audio or video; for production, preprocess images with OCR or vision models if using Jev.
  2. Do you need self-hosted open weights? Clef publishes Apache 2.0 on Hugging Face. Jev: closed. Or see local OSS alternatives.
  3. Are you already on Cloudflare Workers? Clef is native; no external API wiring needed.
  4. Do you want lowest per-token cost? Clef-flash $0.038/M (24k hosted context), Perplexity $0.04/M, Jev $0.042/M, Clef-omni $0.15/M, Clef $0.24/M (as of 15 October 2026; verify current pricing before hard-coding).
  5. Do you want proof builds? Jev has madewithjev catalog with published figures. Clef/Perplexity: vendor benchmarks only as of early October 2026.

For detailed comparison: Jev vs Cloudflare Clef

Try it today

Next steps

Common questions

What are alternatives to Cloudflare Clef?
TypeSafe Jev is the primary alternative with documented GA availability, same question types (Choice/Score/Noul), and published pricing. Perplexity Decisions and OpenAI Decisions (public beta) are peers in the same category. For self-hosted: see open-source alternatives (Laya, Julia 1, kev).
Is TypeSafe Jev a Cloudflare Clef alternative?
Yes. Both are System One decision models handling Choice/Score/Noul question types. Clef has vision, open weights, and audio and video through Clef-omni. Jev has documented GA pricing ($0.042/M input) and, since 14 October 2026, PNG and JPEG image input in preview. On price, Jev sits between Clef-flash ($0.038/M) and Clef-omni ($0.15/M) and Clef ($0.24/M).
Why use Jev instead of Clef?
Use Jev when you want a low per-token cost with a 64k context per request ($0.042/M, below Clef and Clef-omni), want to avoid Workers AI billing, or need builds proven on madewithjev. Use Clef-flash when short inputs fit its 24k hosted context and you want the lowest price ($0.038/M). Use Clef-omni when you need audio or video, and Clef when you need production vision, open weights or Cloudflare Workers.
Can I self-host Jev like Clef?
No: TypeSafe has not released Jev weights. Clef publishes Apache 2.0 weights on Hugging Face. For local alternatives to both, see open-source decision models: Laya (421M), Julia 1 (144M), kev and adapters.
Does Clef work outside Cloudflare Workers?
Yes, via the open weights on Hugging Face. Run it anywhere that can serve the model. Workers AI is the hosted path; the weights are the self-hosted path.

Made with Jev is independent and not affiliated with TypeSafe AI. Every figure on this page is the one its author published, linked to where it can be checked.