Parameters
27.0B
Context
262,144 tokens
native
Modalities
text, image, video text
input → output
Benchmarks
#44/134
better than 67% of all scored models
Size class
#8/21
better than 62% of similar-size models
24h trend
+0.6%
Interest delta, 24h
License
Apache License 2.0
About
Clef is a 27B multimodal decision model from Cloudflare, released on October 1, 2026 under the Apache-2.0 license. It turns a state plus a schema of typed questions into decisions: Clef reads the state as text, JSON, images, or video and returns a probability for every allowed option of every question in a single forward pass - there is no free-form text generation and no output parsing. The Clef API is fully compatible with Jev and SystemOne, and Clef is post-trained from Qwen/Qwen3.8-27B. A smaller, faster variant, Clef-Flash, is also available.
Architecture
Clef consists of the Qwen3.8-27B backbone with its vision encoder (stored as standard sharded safetensors, 27B parameters, BF16) plus a joint schema head - a small transformer head that reads the backbone's final hidden states, routes evidence from the state to each question, and scores all options of all questions jointly in a single forward pass. The output is one logit per allowed option for each question; a softmax per question gives probabilities. Qwen3.8-27B itself is a hybrid linear attention transformer: 64 layers in a 3:1 pattern of 48 linear-attention layers and 16 full-attention layers (gated linear attention with 16 key / 48 value heads at head_dim 128, GQA full attention with 24 query heads / 4 KV heads at head_dim 256, partial rotary factor 0.25, interleaved MRoPE with theta 1e7 and mrope_section [11, 11, 10]), hidden size 5120, intermediate size 17408, vocab 248320, native context 262144 tokens.
Input format and typed questions
A record contains a state (any string or JSON value), optional images/video frames, and a questions mapping. Each typed question declares its allowed answers up front via type (noul true/false, choice named options, or score ordered options), optional instructions, and criteria. encode_record bounds the input at a default max_length of 16,384 tokens (with max_state_tokens), and text-only and multimodal records can be mixed in one batch. The systemone function mirrors the Jev/SystemOne POST /v1/systemone endpoint and returns answers with choice/confidence/probabilities, expected score/legend/probabilities, or the probability of true for noul.
Training and deployment
Clef is post-trained from Qwen3.8-27B (card tag post-train), with the joint schema head trained alongside it; the card documents no post-training data or hyperparameters. Tested with torch 2.11 and transformers 5.10.2 on a single H200 (pillow needed for images/videos). Under the hood it leads Cloudflare's internal Decision Index 0.2.1 suite on routing/classification rows (ToolRet 69.2 nDCG@10, BANKING77 94.2 macro-F1, CLINC150+OOS 97.4, GSM8K 80.8, CRUXEval 86.7, RAGTruth 79.4) with a median latency of 209.3 ms, and leads four Typesafe Evals business workflows on invoice processing and security incidents.
Training Data Post-trained from Qwen/Qwen3.8-27B (with its vision encoder); no pretraining/post-training data recipe documented on the card.
Benchmark Scores
| Benchmark | Score | Date | Standing |
|---|---|---|---|
|
FinEntity
domain_finance
|
96.20
|
— |
#2/2
better than 0%
|
|
Habermas Machine
reasoning
|
68.70
|
— |
#2/2
better than 0%
|
|
HoVer
knowledge
|
100.00%
|
— |
#1/2
better than 100%
|
|
When2Call MCQ
agentic
|
100.00%
|
— |
#1/2
better than 100%
|
|
HellaSwag
reasoning
|
98.90%
|
— |
#2/7
better than 83%
|
|
ARC-c
reasoning
|
99.00%
|
— |
#2/6
better than 80%
|
|
BRIGHT
agentic
|
100.00%
|
— |
#1/2
better than 100%
|
|
BIG-Bench Hard
reasoning
|
77.84%
|
— |
#4/8
better than 57%
|
|
CRUXEval
code
|
100.00%
|
— |
#1/2
better than 100%
|
|
VAST
vision_language
|
100.00%
|
— |
#1/2
better than 100%
|
|
PhishNChips
cybersecurity
|
100.00%
|
— |
#1/2
better than 100%
|
|
New Yorker
composite
|
100.00%
|
— |
#1/2
better than 100%
|
|
BANKING77
domain_finance
|
100.00%
|
— |
#1/2
better than 100%
|
|
ForecastBench
reasoning
|
100.00%
|
— |
#2/2
better than 0%
|
|
GPQA Diamond
stem_reasoning
|
12.46%
|
— |
#79/83
better than 5%
|
|
POP909-CL
general_capabilities
|
100.00%
|
— |
#1/2
better than 100%
|
|
NLI4CT
document_understanding
|
100.00%
|
— |
#1/2
better than 100%
|
|
MMLU-Pro
knowledge
|
19.03%
|
— |
#40/48
better than 17%
|
|
CLINC150+OOS
classification
|
100.00%
|
— |
#1/2
better than 100%
|
|
cfcolor
vision_language
|
100.00%
|
— |
#1/2
better than 100%
|
|
BPoMP
multilingual
|
100.00%
|
— |
#1/2
better than 100%
|
|
ANLI
reasoning
|
100.00%
|
— |
#1/2
better than 100%
|
|
Amazon ESCI
domain_finance
|
100.00%
|
— |
#1/2
better than 100%
|
|
ChessBench
reasoning
|
100.00%
|
— |
#1/2
better than 100%
|
|
ToolRet
agentic
|
100.00%
|
— |
#1/2
better than 100%
|
|
SGD/SGD-X
instruction_following
|
100.00%
|
— |
#1/2
better than 100%
|
|
RAGTruth
safety
|
100.00%
|
— |
#1/2
better than 100%
|
|
ACOS
general_capabilities
|
100.00%
|
— |
#1/2
better than 100%
|
|
WinoGrande
reasoning
|
89.82%
|
— |
#2/6
better than 80%
|
|
GSM8K
math
|
72.78%
|
— |
#3/6
better than 60%
|
|
API-Bank
agentic
|
91.90
|
— |
#2/2
better than 0%
|
|
Humicroedit
reasoning
|
66.70
|
— |
#2/2
better than 0%
|
|
ContractNLI
document_understanding
|
81.40
|
— |
#2/2
better than 0%
|
|
RouterBench
general_agent
|
79.70
|
— |
#2/2
better than 0%
|
|
CLadder
stem_reasoning
|
94.00
|
— |
#2/2
better than 0%
|
|
SATA-Bench
general_agent
|
33.80
|
— |
#2/2
better than 0%
|
|
MuSR
reasoning
|
83.73%
|
— |
#2/3
better than 50%
|
|
ARC-e
reasoning
|
98.11%
|
— |
#2/6
better than 80%
|
|
Home appliance simulator
general_agent
|
83.00
|
— |
#2/2
better than 0%
|
|
MMLU
knowledge
|
91.10%
|
— |
#7/17
better than 62%
|
What is a good FinEntity score? What is a good Habermas Machine score? What is a good HoVer score?
Hardware Requirements
Estimated VRAM per quantization at 4,096 tokens context
| Quantization | Weights | KV Cache | Total VRAM | On your hardware |
|---|---|---|---|---|
| FP16 — 16-bit | 54.0 GB | 4.2 GB | 60.9 GB | Which GPUs can run it? Calculate on your GPU |
| INT8 — 8-bit | 27.0 GB | 2.1 GB | 30.4 GB | Which GPUs can run it? Calculate on your GPU |
| INT4 — 4-bit | 13.5 GB | 1.0 GB | 15.2 GB | Which GPUs can run it? Calculate on your GPU |
| FP4 — 4-bit | 13.5 GB | 1.0 GB | 15.2 GB | Which GPUs can run it? Calculate on your GPU |
Consumer GPUs that fit this model with a single card (INT4):
Architecture
hybrid linear attention transformer (Qwen3.8-27B backbone + joint schema transformer head)
- Parameters
- 27B
- Layers
- 64 48× Linear / Recurrent + 16× Full Attention
- Attention
- Hybrid 24:4 Hybrid Attention
- FFN dim
- 17408
- Hidden
- 5120
- Head dim
- 256
- Vocabulary
- 248K
- Context
- 262K
Source: Hugging Face config.json · Qwen3_5ForConditionalGeneration · exact layer pattern · model repo
Small transformer head stored separately (joint_head.safetensors, joint_head_config.json). It reads the backbone final hidden states, routes evidence from the state to each question in the typed schema, and scores all options of all questions jointly in a single forward pass (one logit per allowed option, softmax per question). This is task-head evidence routing, not MoE expert routing.
Training Pipeline
-
1
other
Post-training from Qwen3.8-27B
The model card states Clef is post-trained from Qwen/Qwen3.8-27B (tagged `post-train`). The released weights include the Qwen3.8-27B backbone with its vision encoder plus a separately stored joint schema head — a small transformer head that reads the backbone's final hidden states, routes evidence from the state to each question, and scores all options of all questions jointly to produce one logit per allowed option. The card does not document the post-training data, method, or any hyperparameters.
Training & Evaluation Datasets
| Name | Role | Size | Modalities | Collection |
|---|---|---|---|---|
| Decision Index 0.2.1 | evaluation | — | text, image, video | — |
| Typesafe Evals | evaluation | 4 end-to-end business workflows | — |
Linked Resources
Trend Analysis
24h Change
+0.6%
7d Change
+1.7%
Current
1,432
downloads_all_time
+1.3%
downloads
+0.8%
likes
+0.2%
No comments yet
Be the first to share your thoughts!