Back to Models
DeepSeek V4 Flash Latest
Deepseektext9.0 / 10 Overall Rating
DeepSeek V4 Flash Latest is an ultra-low-cost, text-only model alias featuring a 1,310,720 token context window and 131,072 max output capacity. Developed by DeepSeek, it targets high-throughput tasks at $0.03 per million input tokens, drastically undercutting GPT-4o and Claude 3.5 Sonnet on price and context capacity.
Context Window
1311K
Knowledge Cutoff
2025
Max Output
16K
License
Proprietary
Where it Excels (Pros)
- Extremely low input output cost
- 1.3M token context window
- 131K max output token capacity
Limitations (Cons)
- No vision or audio support
- Lower reasoning depth than Sonnet
Benchmark Breakdown
Reasoning (MMLU)82%
Coding (HumanEval)78%
Mathematics (MATH)76%
Long-Context Retrieval95%
Pricing Matrix
Input cost / 1M tokens$0.03
Output cost / 1M tokens$0.10
Compare DeepSeek V4 Flash Latest with other frontier models: