DeepSeek-V4-Flash-DSpark
Support summary
DeepSeek-V4-Flash-DSpark is a checkpoint published by DeepSeek, built on the DeepseekV4ForCausalLM architecture. Official recipes or cookbooks cover vLLM, without pinning a stable release.
- Documented hardware
- NVIDIA / CUDA and AMD / ROCm
- Documented quantization
- FP8
- Configured context
- 1,048,576 tokens
- License
- mit
Hardware, quantization and features count only statements about this checkpoint; engine-wide capabilities are listed with each release.
Hugging Face model card
- Publisher
- DeepSeek
- Task
- text generation
- License
- mit
- Library
- transformers
- Architecture
- DeepseekV4ForCausalLM
- Weight tensor elements
- 165.3B
- Tensor types
- BF16, I64, F32, F8_E8M0, F8_E4M3, I8
- Quantization
- fp8
- Configured context
- 1,048,576 tokens
- Weight access
- Public
- Repository created
- 2026-06-27
Engine support
3 engines
| Engine | Release | Checkpoint support | Evidence scope | Last known supported |
|---|---|---|---|---|
| vLLM | Version not pinnedRecipe / cookbook | Documented recipeSource: vllm · models/deepseek-ai/DeepSeek-V4-Flash.yamlReviewed | No supported release indexed | |
Features, dependencies and evidence for vLLM 0.28.0 | ||||
| SGLang | 0.5.19Newest indexed | No matching evidence curated | No supported release indexed | |
Features, dependencies and evidence for SGLang 0.5.19 | ||||
| TensorRT-LLM | 1.2.1Newest indexed | No matching evidence curated | No supported release indexed | |
Features, dependencies and evidence for TensorRT-LLM 1.2.1 | ||||