Cloudflare Clef

What hackers pick instead

• Jev by Typesafe AIA cheap, low-latency general classifier you redefine with a prompt instead of retraining, though critics call it overhyped and little more than standard LLM tech.392recommend
• LayaOpen-weight, self-hostable decision model that people run locally and fine-tune for specific classification tasks, with fast responses. Common caveat: weak out-of-box accuracy and a 1024-token context limit.44recommend
• OpenJevFast, cheap structured-output model that cannot emit tokens outside a given set. Critics doubt it differs from other small LLMs, and the naming is confusing.21recommend
• Typesafe AIPeople recommend it as a cheap API model that handles new classification problems without fine-tuning. Critics say it is over-hyped and not production-ready without fine-tuning.16recommend
• Strands Decider 2BA small open-source decision model that people run locally for filtering and automations. Critics note a bias toward the first option listed.7recommend

Every quote

The 32 most detailed of 36 comments. Each links to the original.

1. The pricing is ... hm more expensive but not at the point I won't give it a try due to the embeded vision encoding

   Mixed · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

2. clef beat Jev pretty handily, I think there was actually 0 instances where clef regressed

   Recommends · Oct 2026 · in “Typesafe AI raises $870M at $7.5B”

3. the value of Clef is mostly that it's open source and I can run it in my infra

   Mixed · Oct 2026 · in “Jev vs. Clef”

4. If CF's benchmark is representative and sufficient, Clef outperforms Jev!

   Recommends · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

5. Pricing is $0.24/million input tokens which is ~6x compared to Jev.

   Mixed · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

6. Because Clef is a larger model, larger context, can process images.

   Recommends · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

7. By itself, it plays in ways you can likely imagine.. not well.

   Mixed · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

8. Times the model sacrificed the individual to save the the five

   Mixed · Oct 2026 · in “Jev meets the Trolley Problem”

9. the first thing I've found local models genuinely useful for

   Recommends · Oct 2026 · in “Bag of Decisions Reranker”

10. it becomes just plug and play. No dependencies needed.

    Recommends · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

11. Clef-flash is at $0.09 which is way more competitive.

    Recommends · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

12. Only Liquid D1 and Clef matched Jev's performance.

    Recommends · Oct 2026 · in “Decisions API is in public beta”

13. Hosted p50: Clef ~850ms, Jev ~110ms

    Mixed · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

14. It allows image input. Nice!

    Recommends · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

15. but it's larger and slower

    Mixed · Oct 2026 · in “Strands Decider 2B: a small, open-source, decision model”

16. I ported a Jev-based prototype i've been working on to Clef, and it ended up being significantly slower and the results were worse.

    Criticizes · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

17. clef came out 5.2x more expensive, a lot slower (p50 of 350ms vs 1.9s) with only marginally better results

    Criticizes · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

18. Clef was 2-3x slower and worse (it caught less hate speech) than Jev.

    Criticizes · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

19. Clef was worse in accuracy, and an order of magnitude slower.

    Criticizes · Oct 2026 · in “Typesafe AI raises $870M at $7.5B”

20. Tested, and it's bad. Slower and wronger than Jev.

    Criticizes · Oct 2026 · in “We’re releasing Jev 1.14.0 along with experimental features for dreamers”

21. One million decisions on Clef cost about $72.

    Criticizes · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

22. it sounds like it is still 10x as slow as Jev

    Criticizes · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

23. it takes 3s which is useless to me

    Criticizes · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

24. Clef-flash: over-escalates

    Criticizes · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

25. Competition in this area is great and kudos for releasing something that we can try out today.

    Recommends · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

26. The pricing is ... hm more expensive but not at the point I won't give it a try

    Mixed · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

27. Clef seems to be a pretty strong attempt at a high difficulty UX.

    Recommends · Aug 2026 · in “Show HN: Clef – spaced repetition for piano over MIDI”

28. What a great application of spaced-repetition!

    Recommends · Aug 2026 · in “Show HN: Clef – spaced repetition for piano over MIDI”

29. Response time of 300ms it's wild!

    Recommends · Oct 2026

30. If I have to gather and tag data to fine-tune Jev, I can probably just train an "old school" classifier model and make it even cheaper, faster, and just as accurate.

    Criticizes · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

31. Characterizing it as "faster than Jev" is very misleading

    Criticizes · Oct 2026 · in “Clef: Open-weight decision models, and new RL fine-tuning platform”

32. Go ask Jev devs how they feel about Clef.

    Criticizes · Oct 2026 · in “AI Makes Me Sad”