Decision modelActive
Clef
Cloudflare
- Released
- October 1, 2026
- Data date
- October 3, 2026
Clef is Cloudflare’s 27-billion-parameter model for typed decisions. You provide state, questions, and possible answers. Choice, Noul, and Score come back as structured probabilities without free-form answer text. The same decision run can take text, JSON, images, and video as input.
The checkpoint is based on Qwen3.8-27B with a vision encoder. Cloudflare specifies a 64k-token context. You can call it through Workers AI as @cf/cloudflare/clef for $0.24 per 1M input tokens or run the Apache-2.0 weights locally. The free neuron allowance is shared across the entire Workers AI account. Clef is a decision component, not a general chat model.
Specifications and access
| Specification | Value and source |
|---|---|
| Model class | Multimodal System One decision model, not a chat modelSource |
| Output schema | Choice, Noul and Score with probabilities, without free-form answer textSource |
| Input | Text, JSON, images, and videoSource |
| Access | Workers AI API and local weightsSource |
| API model ID | Workers AI: @cf/cloudflare/clefSource |
| Checkpoint | Cloudflare/clefSource |
| Base model | Qwen/Qwen3.8-27B with vision encoderSource |
| Context window | 64k tokens (provider specification)Source |
| License | Apache-2.0Source |
Page 1 of 4
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
Decision Index 0.2.1: BFCL (case exact accuracy) (Show measurement, test conditions, and source)
- Source value
- 98.5
- Score
- 98.5
- Metric
- BFCL (case exact accuracy)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: ToolRet (nDCG@10) (Show measurement, test conditions, and source)
- Source value
- 69.2
- Score
- 69.2
- Metric
- ToolRet (nDCG@10)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: API-Bank (accuracy) (Show measurement, test conditions, and source)
- Source value
- 91.9
- Score
- 91.9
- Metric
- API-Bank (accuracy)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: BANKING77 (macro-F1) (Show measurement, test conditions, and source)
- Source value
- 94.2
- Score
- 94.2
- Metric
- BANKING77 (macro-F1)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: CLINC150+OOS (macro-F1) (Show measurement, test conditions, and source)
- Source value
- 97.4
- Score
- 97.4
- Metric
- CLINC150+OOS (macro-F1)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: RouterBench (selected quality) (Show measurement, test conditions, and source)
- Source value
- 79.7
- Score
- 79.7
- Metric
- RouterBench (selected quality)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: Home appliance simulator (case exact accuracy) (Show measurement, test conditions, and source)
- Source value
- 83
- Score
- 83
- Metric
- Home appliance simulator (case exact accuracy)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: SGD/SGD-X (macro-F1) (Show measurement, test conditions, and source)
- Source value
- 43.8
- Score
- 43.8
- Metric
- SGD/SGD-X (macro-F1)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: ContractNLI (macro-F1) (Show measurement, test conditions, and source)
- Source value
- 81.4
- Score
- 81.4
- Metric
- ContractNLI (macro-F1)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: ANLI (macro-F1) (Show measurement, test conditions, and source)
- Source value
- 69.8
- Score
- 69.8
- Metric
- ANLI (macro-F1)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: BPoMP (accuracy) (Show measurement, test conditions, and source)
- Source value
- 96.9
- Score
- 96.9
- Metric
- BPoMP (accuracy)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Decision Index 0.2.1: Humicroedit (accuracy) (Show measurement, test conditions, and source)
- Source value
- 66.7
- Score
- 66.7
- Metric
- Humicroedit (accuracy)
- Unit
- %
- Benchmark version
- 0.2.1
- Category
- Decision Index 0.2.1
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cloudflare Clef model card
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Run=Cloudflare internal evaluation; Benchmark version=0.2.1; Request latency=no
Page 1 of 4
Published prices
Prices apply to the stated unit. Resolution, output length, and provider can change the cost.
Sources and data date
Every statement links to its underlying documentation or leaderboard.
| Type | Evidence and data date |
|---|---|
| Research status | Research date October 4, 2026. 1 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Unresolved | forecastbench-scale Original wording ForecastBench prints Brier values 13.9 and 10.6 without a stated scale; the raw values remain unchanged and are not divided by 100. |
| Additional source | Cloudflare announcement (retrieved October 3, 2026) · Cloudflare Clef announcement · Editorial description reviewed October 3, 2026 |
| Additional source | Clef model card (retrieved October 3, 2026; October 4, 2026) · Clef model card · Editorial description reviewed October 3, 2026 |
| Additional source | Workers AI pricing (retrieved October 3, 2026) · Workers AI pricing · Editorial description reviewed October 3, 2026 |