Skip to main content
BenchLM
Data

Cloudflare Decision Index 0.2.1: ForecastBench

We show this table for reference; we do not rank on it.

Published Cloudflare and community decision-model configurations.

Evaluation results published by Cloudflare and upstream community contributors.

Brier (lower is better) on Cloudflare Decision Index 0.2.1: ForecastBench — October 1, 2026 capture

We mirror the published brier (lower is better) view for Cloudflare Decision Index 0.2.1: ForecastBench. Clef-flash leads the public snapshot at 0.106, followed by Clef (0.139) and Eikos-27B [FP8] (0.171). We do not use these results to rank models overall.

73 configurationsDecision ModelsDated mixed-source configurationsDisplay onlyUpdated October 1, 2026 capture

Brier (lower is better) table (73 configurations)

Brier (lower is better) results for Cloudflare Decision Index 0.2.1: ForecastBench
RankModel / configurationScoreParameters (B)Open / closed
1Clef-flashCloudflareCloudflare self-reported; Routing head + LoRA; 36/38 index tasks; median 38.8 ms; p95 122.4 ms
0.106
Not reportedUnknown
2ClefCloudflareCloudflare self-reported; Routing head + LoRA; 36/38 index tasks; median 209.3 ms; p95 238.6 ms
0.139
Not reportedUnknown
3Eikos-27B [FP8]Creator not reportedUpstream community result; LoRA; 38/38 index tasks; median 129.7 ms; p95 509.8 ms
0.171
Not reportedUnknown
4JevTypeSafeUpstream community result; Hosted API (closed); 38/38 index tasks; median 524.1 ms; p95 536 ms
0.174
Not reportedUnknown
5Kev 9BCreator not reportedUpstream community result; LoRA + head; 38/38 index tasks; median 51.4 ms; p95 187.9 ms
0.176
Not reportedUnknown
6Jebadiah 27BCreator not reportedUpstream community result; LoRA; 38/38 index tasks; median 110.4 ms; p95 591.1 ms
0.178
Not reportedUnknown
7JPT-9BCreator not reportedUpstream community result; LoRA; 38/38 index tasks; median 148.6 ms; p95 461.9 ms
0.181
Not reportedUnknown
8Kev 4BCreator not reportedUpstream community result; LoRA + head; 38/38 index tasks; median 52.1 ms; p95 141.1 ms
0.186
Not reportedUnknown
9Decider chat · Qwen3.6-27BCreator not reportedUpstream community result; inference technique; 38/38 index tasks; median 83.6 ms; p95 965.6 ms
0.186
Not reportedUnknown
10JPT-4BCreator not reportedUpstream community result; LoRA; 38/38 index tasks; median 139.1 ms; p95 405.7 ms
0.189
Not reportedUnknown
11Decision 1.0 LuxCreator not reportedUpstream community result; head / adapter; 38/38 index tasks; median 51 ms; p95 347.5 ms
0.191
Not reportedUnknown
12reflex 27B [FP8 · wide choice]Creator not reportedUpstream community result; inference technique; 38/38 index tasks; median 108.3 ms; p95 1239.7 ms
0.192
Not reportedUnknown
13Winnow-12B [Q8_0]Creator not reportedUpstream community result; LoRA; 38/38 index tasks; median 72.5 ms; p95 356.2 ms
0.193
Not reportedUnknown
14AutoJev-27BCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 101.4 ms; p95 1052.5 ms
0.195
Not reportedUnknown
15JevK5Creator not reportedUpstream community result; LoRA; 38/38 index tasks; median 22 ms; p95 226.3 ms
0.199
Not reportedUnknown
16Tev1-4B-experimentalCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 35.8 ms; p95 166 ms
0.200
Not reportedUnknown
17Decider 35B-A3B [NVFP4]Creator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 101.4 ms; p95 210.9 ms
0.201
Not reportedUnknown
18XorCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 139.9 ms; p95 419.9 ms
0.201
Not reportedUnknown
19Jev-OmniCreator not reportedUpstream community result; LoRA + head; 38/38 index tasks; median 54.9 ms; p95 627.9 ms
0.202
Not reportedUnknown
20Decider chat · Gemma-4-31BCreator not reportedUpstream community result; inference technique; 38/38 index tasks; median 108.5 ms; p95 1114.7 ms
0.202
Not reportedUnknown
21Surogate Rune 26B-A4B v3 [bf16]Creator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 120.5 ms; p95 682.8 ms
0.203
Not reportedUnknown
22Decision 1.0 NoxCreator not reportedUpstream community result; head / adapter; 38/38 index tasks; median 53.2 ms; p95 242.1 ms
0.209
Not reportedUnknown
23Intern-Decision-4BCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 44.2 ms; p95 115.4 ms
0.211
Not reportedUnknown
24NeoHorse-Jev-4BCreator not reportedUpstream community result; head / adapter; 38/38 index tasks; median 49 ms; p95 273.9 ms
0.211
Not reportedUnknown
25JevfireCreator not reportedUpstream community result; inference technique; 38/38 index tasks; median 89.6 ms; p95 447.5 ms
0.213
Not reportedUnknown
26simple-jev · Qwen3.8-27B (featherless)Creator not reportedUpstream community result; inference technique; 38/38 index tasks; median 373 ms; p95 2796.8 ms
0.213
Not reportedUnknown
27mmastrac diffusiongemma (vLLM PR 57250)Creator not reportedUpstream community result; inference technique; 38/38 index tasks; median 126.7 ms; p95 274.6 ms
0.214
Not reportedUnknown
28levCreator not reportedUpstream community result; LoRA + head; 38/38 index tasks; median 70.8 ms; p95 687.7 ms
0.216
Not reportedUnknown
29open-jev (pngwn)Creator not reportedUpstream community result; LoRA; 38/38 index tasks; median 112.3 ms; p95 724.1 ms
0.218
Not reportedUnknown
30Jet v6.2Creator not reportedUpstream community result; LoRA; 38/38 index tasks; median 46.8 ms; p95 371.1 ms
0.223
Not reportedUnknown
31djevCreator not reportedUpstream community result; inference technique; 38/38 index tasks; median 84.4 ms; p95 211.2 ms
0.223
Not reportedUnknown
32Bespoke Nimble 9B v2Creator not reportedUpstream community result; LoRA; 38/38 index tasks; median 77.3 ms; p95 621.4 ms
0.224
Not reportedUnknown
33Hopper (G) 1.2Creator not reportedUpstream community result; LoRA; 38/38 index tasks; median 23.3 ms; p95 238.2 ms
0.224
Not reportedUnknown
34JobeCreator not reportedUpstream community result; inference technique; 38/38 index tasks; median 53.5 ms; p95 2910.3 ms
0.225
Not reportedUnknown
35SemIfCreator not reportedUpstream community result; inference technique; 38/38 index tasks; median 113.1 ms; p95 214.6 ms
0.225
Not reportedUnknown
36VerdictCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 11.4 ms; p95 60.1 ms
0.227
Not reportedUnknown
37JPT-0.8BCreator not reportedUpstream community result; LoRA; 38/38 index tasks; median 117.5 ms; p95 341.2 ms
0.235
Not reportedUnknown
38razorback16 openjev diffusiongemma (NVFP4, vLLM) [one read, NVFP4]Creator not reportedUpstream community result; inference technique; 38/38 index tasks; median 37.7 ms; p95 117.8 ms
0.236
Not reportedUnknown
39Decision 1.0 EosCreator not reportedUpstream community result; head / adapter; 38/38 index tasks; median 39.7 ms; p95 84.3 ms
0.236
Not reportedUnknown
40Kev 0.8BCreator not reportedUpstream community result; LoRA + head; 38/38 index tasks; median 41.6 ms; p95 95.9 ms
0.237
Not reportedUnknown
41Intern-Decision-0.8BCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 33.7 ms; p95 39.2 ms
0.238
Not reportedUnknown
42Solomon v1.1Creator not reportedUpstream community result; head / adapter; 38/38 index tasks; median 223.9 ms; p95 1176.8 ms
0.238
Not reportedUnknown
43Intern-Decision-2BCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 33.5 ms; p95 58 ms
0.239
Not reportedUnknown
44GLiNER2.5-DecideCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 23.3 ms; p95 228.6 ms
0.239
Not reportedUnknown
45Lumma-Fev-0.1BCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 20.9 ms; p95 112.5 ms
0.240
Not reportedUnknown
46Winnow-E4B [Q8_0]Creator not reportedUpstream community result; LoRA; 38/38 index tasks; median 45 ms; p95 188.3 ms
0.245
Not reportedUnknown
47Tev1-0.8B-experimentalCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 28.5 ms; p95 133.9 ms
0.247
Not reportedUnknown
48Qwen-2.5-1B-RLCDCreator not reportedUpstream community result; inference technique; 38/38 index tasks; median 43.8 ms; p95 75.2 ms
0.247
Not reportedUnknown
49Lumma-Fev-0.6BCreator not reportedUpstream community result; LoRA; 38/38 index tasks; median 21.3 ms; p95 219.7 ms
0.255
Not reportedUnknown
50Metask-Jev-4BCreator not reportedUpstream community result; LoRA; 38/38 index tasks; median 57.3 ms; p95 369.4 ms
0.270
Not reportedUnknown
51Decider 2B [FP8]Creator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 8.1 ms; p95 134.9 ms
0.271
Not reportedUnknown
52CLM-v0.1-8BCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 46.8 ms; p95 130.9 ms
0.281
Not reportedUnknown
53Bosun v3.1 0.6BCreator not reportedUpstream community result; LoRA + head; 38/38 index tasks; median 38.7 ms; p95 107.3 ms
0.289
Not reportedUnknown
54jeffCreator not reportedUpstream community result; inference technique; 38/38 index tasks; median 21.8 ms; p95 67.6 ms
0.291
Not reportedUnknown
55Decider 4BCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 12.6 ms; p95 188.5 ms
0.294
Not reportedUnknown
56JoshuaSP diffusiongemma (open-jev)Creator not reportedUpstream community result; inference technique; 38/38 index tasks; median 260.4 ms; p95 537.3 ms
0.296
Not reportedUnknown
57LFM2.5-2.6B-RLCDCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 39.3 ms; p95 95.4 ms
0.303
Not reportedUnknown
58this-that 1.2Creator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 44.3 ms; p95 57.2 ms
0.327
Not reportedUnknown
59Bosun v3.1 1.7BCreator not reportedUpstream community result; LoRA + head; 38/38 index tasks; median 40.3 ms; p95 173.5 ms
0.331
Not reportedUnknown
60GLiNER 2.5 multilingualCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 14 ms; p95 92.4 ms
0.334
Not reportedUnknown
61Decision 1.0 SolCreator not reportedUpstream community result; head / adapter; 38/38 index tasks; median 38.1 ms; p95 101 ms
0.336
Not reportedUnknown
62MoJevCreator not reportedUpstream community result; head / adapter; 38/38 index tasks; median 46.4 ms; p95 55.1 ms
0.346
Not reportedUnknown
63LavoirCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 20 ms; p95 49.5 ms
0.362
Not reportedUnknown
64system-one-gemmaCreator not reportedUpstream community result; LoRA + head; 38/38 index tasks; median 31.9 ms; p95 748 ms
0.372
Not reportedUnknown
65openvonsCreator not reportedUpstream community result; inference technique; 38/38 index tasks; median 21.2 ms; p95 113.3 ms
0.387
Not reportedUnknown
66LayaCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 5.8 ms; p95 222.5 ms
0.411
Not reportedUnknown
67Julia 1Creator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 5.8 ms; p95 15.3 ms
0.411
Not reportedUnknown
68Decision 1.0 KaiCreator not reportedUpstream community result; head / adapter; 38/38 index tasks; median 30.4 ms; p95 106.8 ms
0.434
Not reportedUnknown
69GLiNER 2.5 baseCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 14.1 ms; p95 62.1 ms
0.466
Not reportedUnknown
70mini-jevCreator not reportedUpstream community result; inference technique; 38/38 index tasks; median 66.9 ms; p95 178 ms
0.474
Not reportedUnknown
71Decision 1.0 LexCreator not reportedUpstream community result; head / adapter; 38/38 index tasks; median 29.8 ms; p95 106 ms
0.477
Not reportedUnknown
72GLiNER 2.5 smallCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 13.6 ms; p95 35.2 ms
0.508
Not reportedUnknown
73LFM2.5-350M-RLCDCreator not reportedUpstream community result; full fine-tune; 38/38 index tasks; median 26.7 ms; p95 174.3 ms
0.622
Not reportedUnknown

How to read this source comparison

The source combines Cloudflare self-reported Clef runs with upstream community configurations. Each row retains its source engine and evidence label; model names alone do not establish equivalent serving settings.

The 0.2.1 composite uses 38 panel benchmarks grouped into five weighted areas. It applies chance correction; ForecastBench uses its Brier transformation. Additional task views retain the source metric and panel-inclusion metadata. The snapshot lists 48 task metrics, while the launch refers to 43 evaluations including latency rows.

Clef and Clef-Flash each cover 36 of 38 index tasks; HLE and iSarcasmEval are missing. Missing values are not inserted as zero. The composite is retained as published without recomputing it or filling gaps.

Community results use the upstream September 28 snapshot, which declares one NVIDIA RTX PRO 6000. Cloudflare run hardware is not established by this artifact. Latency comparisons across these configurations do not isolate model speed. The October 1 date marks capture, not evaluation.

These views remain separate from JevBench and canonical provider benchmark results. They do not enter overall or category rankings. Equal scores do not establish statistical ties.

The published Cloudflare Decision Index 0.2.1: ForecastBench snapshot places Clef-flash first at 0.106. The third row is 0.065 score units higher. The broader top-10 range is 0.083 score units, so many of the published results sit in a relatively narrow band.

73 configurations are shown for Cloudflare Decision Index 0.2.1: ForecastBench. The benchmark falls in the Decision Models category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. Cloudflare Decision Index 0.2.1: ForecastBench is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About Cloudflare Decision Index 0.2.1: ForecastBench

Year

2026

Freshness and provenance

Version

Cloudflare comparison on Decision Index 0.2.1

Refresh cadence

Reviewed dated source capture

Staleness state

Dated mixed-source configurations

Question availability

Upstream tasks and terms; aggregate scores only

Dated mixed-source configurationsDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Questions

What does Cloudflare Decision Index 0.2.1: ForecastBench measure?

Published Cloudflare and community decision-model configurations.

Which model leads the published Cloudflare Decision Index 0.2.1: ForecastBench snapshot?

Clef-flash currently leads the published Cloudflare Decision Index 0.2.1: ForecastBench snapshot with 0.106 brier (lower is better). BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on Cloudflare Decision Index 0.2.1: ForecastBench?

The October 1, 2026 capture snapshot contains 73 AI models.

Last updated: October 1, 2026 capture · mirrored from the public benchmark leaderboard

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 5,600+ readers.

One email each week. Unsubscribe anytime.