clef
Q4 GGUF download: ggml-org/Clef-GGUF - 19.54 GB on disk, read 2026-11-21. VRAM figures above are estimates; this is the actual file size.
Will it run on your hardware?
Pick your GPU memory - see which quantizations fit, and the cheapest card for the rest
Need an exact figure for your context length? Use the VRAM calculator.
Run it locally
Copy-paste - running in under a minute
vllm serve Cloudflare/clefNew to this? Start with Ollama · serve to many users with vLLM.
Deep dive
Notes, sources, and the full write-up
clef is an open-weight vision-language model published by Cloudflare on Hugging Face in September 2026, with 27.4B parameters under the apache-2.0 license. It is built on Qwen/Qwen3.8-27B.
Among the 72 vision-language models in this directory it ranks 42nd by monthly Hugging Face downloads, and 31 of the others are larger.
HardwareLink to this section
Unquantized FP16 weights take about 55 GB; 8-bit brings that to ~29 GB and 4-bit (Q4_K_M) to ~16 GB, before activations and any context or image buffers. Even at 4-bit it needs more than 12 GB, so plan on a 24 GB setup. Check exact numbers for your card with the VRAM calculator.
Similar vision-language modelsLink to this section
- Qwen3.8 27B - 27.8B, 6,285,486 downloads/month
- Gemma 4 31B - 32.7B, 6,264,804 downloads/month
- Gemma 4 26B A4B - 26.5B, 5,022,424 downloads/month
- Qwen3.6 27B - 28B, 2,479,223 downloads/month
- AutoTrust-Decider-Pro-Vision-27B - 27.8B, 1,091,125 downloads/month
PopularityLink to this section
clef was downloaded 78,027 times on Hugging Face in the last month (100,039 all-time) and has 2,377 likes.
Run it locallyLink to this section
Weights and usage instructions are on the Cloudflare/clef model card; it loads with Hugging Face Transformers.
Local runtimes that fit this model:
- vllm - high-throughput serving on NVIDIA GPUs
- transformers - the reference Python library most model cards target
Frequently asked
Quick answers to common questions
How much VRAM does clef need?
clef with 27.4B parameters needs approximately 16 GB at Q4_K_M quantization. Use our VRAM calculator for an exact estimate.
Is clef better than other Cloudflare models?
clef has 27.4B parameters with 262,144 context - a strong choice for vision, long-context.
What license is clef under?
clef is released under the apache-2.0 license, making it suitable for most commercial and personal projects.
What hardware runs clef well?
With 27.4B parameters, clef requires adequate VRAM. High-end GPUs like the RTX 4090 (24GB), RTX 5090 (32GB), or Mac Studio with unified memory are good options. Check our hardware directory for specific recommendations.
What is the best quantization for clef?
Q4_K_M is the recommended sweet spot - ~98% of FP16 quality at ~27% of the size. Q5_K_M (~19 GB) is an option if you have spare VRAM. Use our VRAM calculator to compare.
How long can clef's context window handle?
clef supports a 262,144-token context window - enough for very long documents, codebases, or multi-turn conversations. Real-world usable context may vary by implementation.
What models compete with clef?
clef competes with other models in its class. Browse our model directory for comparisons, benchmarks, and community reviews to find the best fit.
Nearby options
Similar models and compatible hardware by spec