Back
DeepSeek: DeepSeek V3
DeepSeek
Input: text
Output: text
Released: Dec 26, 2024•Updated: Mar 28, 2025
DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.
For model details, please visit the DeepSeek-V3 repo for more information, or see the launch announcement.
163,840 Token Context
Process and analyze large documents and conversations.
Advanced Coding
Improved capabilities in front-end development and full-stack updates.
Agentic Workflows
Autonomously navigate multi-step processes with improved reliability.
Available On
| Provider | Model ID | Context | Max Output | Input Cost | Output Cost | Throughput | Latency |
|---|---|---|---|---|---|---|---|
| DeepInfra | deepInfra | 164K | 164K | $0.38/M | $0.89/M | 22.1 t/s | 669 ms |
| Novita | novitaAi | 64K | 16K | $0.40/M | $1.30/M | 25.7 t/s | 1116 ms |
| Nebius | nebiusAiStudio | 164K | - | $0.50/M | $1.50/M | 22.5 t/s | 383 ms |
| Fireworks | fireworks | 131K | - | $0.90/M | $0.90/M | 50.8 t/s | 1799 ms |
Standard Pricing
Input Tokens
$0.00000038
per 1K tokens
Output Tokens
$0.00000089
per 1K tokens