DeepSeek-R1-Zero
transformerssafetensorsdeepseek_v3text-generationconversationalcustom_codearxiv:2501.12948license:mittext-generation-inferenceendpoints_compatiblefp8region:us
Contexto
131,072 tokens
Output máx
8,192 tokens
Lançado
20/01/2025
Provedores
1
Overview
DeepSeek-R1-Zero is a large language model developed by DeepSeek. It supports 131,072 tokens context and is available through 1 API provider(s).
Key Highlights
- Unknown / Not disclosed
Brand
DeepSeek
Organization
País: China
View all models, pricing and details from DeepSeek.
Specifications
Basic Information
| Modelo | DeepSeek-R1-Zero |
| Marca | DeepSeek |
| Version | 1 |
| Model Type | transformers |
| Status | stable |
| Lançado | 20/01/2025 |
| Popularidade | 60420 |
Architecture
| Parameter Size | 684.5B |
| Architecture Type | deepseek_v3 |
| Attention Mechanism | causal-moe-mla |
| Tokenizer | Unknown / Not disclosed |
| Vocabulary Size | Unknown / Not disclosed |
Context & I/O
| Contexto | 131,072 tokens |
| Output máx | 8,192 tokens |
| Modalities | Unknown / Not disclosed |
| Languages | Unknown / Not disclosed |
Capabilities
| Vision | No |
| Audio | No |
| Reasoning | No |
| Function Call | No |
| MCP | No |
| JSON Mode | No |
| Embedding | No |
Training Data
| Training Dataset | Unknown / Not disclosed |
| Estimated Data Size | Unknown / Not disclosed |
| Knowledge Cutoff | Unknown / Not disclosed |
Inference & Deployment
| API Access | Yes |
| VRAM Estimate | Unknown / Not disclosed |
| Quantization Support | Unknown / Not disclosed |
| Hardware Requirements | Unknown / Not disclosed |
Safety & Compliance
| Safety Evaluations | Unknown / Not disclosed |
| Content Policy | Unknown / Not disclosed |
| Region Restrictions | Unknown / Not disclosed |
License & Pricing
| License | mit |
| Commercial Use | Unknown / Not disclosed |
| Open Source | Yes |
Provider & Pricing
| Provedor | Input | Output | Cache | Batch | Website |
|---|---|---|---|---|---|
| deepseek-ai | 0,000 €/1M | 0,000 €/1M | - | - | Visitar website |
Alterações de preço
Nenhuma alteração de preço registrada.
Benchmarks
No benchmark data available.
Changelog
- Page created / data imported23/09/2026
- Last updated23/09/2026
FAQ
What is DeepSeek-R1-Zero?
DeepSeek-R1-Zero is a large language model developed by DeepSeek.
Who developed DeepSeek-R1-Zero?
DeepSeek-R1-Zero was developed by DeepSeek. You can learn more on the brand page or official website.
What is the context length of DeepSeek-R1-Zero?
DeepSeek-R1-Zero supports up to 131,072 tokens of context.
How much does DeepSeek-R1-Zero API cost?
API pricing for DeepSeek-R1-Zero starts at 0,000 €/1M per 1M input tokens and 0,000 €/1M per 1M output tokens. See the pricing table above for provider-specific rates.