05/28 updated checkpoint of Deepseek R1. Its overall performance is now approaching that of leading models, such as O3 and Gemini 2.5 Pro. Compared to the previous version, the upgraded model shows significant improvements in handling complex reasoning tasks, and this version also offers a reduced hallucination rate, enhanced support for function calling, and better experience for vibe coding. Note that fine-tuning for this model is only available through contacting fireworks at https://fireworks.ai/company/contact-us.
Fine-tuningDocs | Deepseek R1 05/28 can be customized with your data to improve responses. Fireworks uses LoRA to efficiently train and deploy your personalized model |
ServerlessDocs | Immediately run model on pre-configured GPUs and pay-per-token |
On-demand DeploymentDocs | On-demand deployments give you dedicated GPUs for Deepseek R1 05/28 using Fireworks' reliable, high-performance system with no rate limits. |
Run queries immediately, pay only for usage
DeepSeek R1-0528 is a large language model developed by DeepSeek AI. This version is a May 28, 2025 update to the DeepSeek R1 model series, offering significant improvements in reasoning, reduced hallucination, enhanced function calling, and stronger performance on tasks like coding and math.
DeepSeek R1-0528 is optimized for:
The model supports a maximum context length of 163.8k tokens.
The default temperature is 0.6.
Fireworks allows a maximum generation length of 64k tokens.
Yes, function calling is supported.
The model has 671B parameters.
Yes, fine-tuning is supported on Fireworks AI using LoRA.
DeepSeek R1-0528 is released under the MIT License, and commercial use and distillation are allowed.
