May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active...
Model details →DeepSeek: R1 0528 vs Meta: Llama 3.1 405B Instruct
DeepSeek: R1 0528 wins on 4 of 6 axes — pricing and capability skew in its favour for most workloads.
The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs. Meta's latest class of model (Llama 3.1) launched with a var…
Model details →Side-by-side comparison
| Capability | DeepSeek: R1 0528 | Meta: Llama 3.1 405B Instruct | Winner |
|---|---|---|---|
| Context window Maximum number of input tokens the model can attend to in a single request. | 164K | 131K | 🏆 DeepSeek: R1 0528 |
| Input price (per 1M) Cost per million input tokens billed by the provider. | $0.50 | $4.00 | 🏆 DeepSeek: R1 0528 |
| Output price (per 1M) Cost per million output tokens billed by the provider. | $2.15 | $4.00 | 🏆 DeepSeek: R1 0528 |
| Tool / function calling First-class support for emitting structured tool calls. | Yes | Yes | Tie |
| Vision input Accepts image inputs alongside text. | — | — | Tie |
| Reasoning mode Internal chain-of-thought / extended-thinking support. | Yes | — | 🏆 DeepSeek: R1 0528 |
Frequently asked questions
Is DeepSeek: R1 0528 better than Meta: Llama 3.1 405B Instruct?
DeepSeek: R1 0528 wins on 4 of 6 axes — pricing and capability skew in its favour for most workloads.
What's the price difference between DeepSeek: R1 0528 and Meta: Llama 3.1 405B Instruct?
Input: $0.50 vs $4.00 per 1M tokens. Output: $2.15 vs $4.00 per 1M tokens.
What context windows do DeepSeek: R1 0528 and Meta: Llama 3.1 405B Instruct support?
DeepSeek: R1 0528 supports up to 164K tokens. Meta: Llama 3.1 405B Instruct supports up to 131K tokens.
Do both DeepSeek: R1 0528 and Meta: Llama 3.1 405B Instruct support tool calling?
DeepSeek: R1 0528: yes. Meta: Llama 3.1 405B Instruct: yes.