DeepSeek V4 Flash (Thinking)
DeepSeek V4 Flash Thinking enables DeepSeek's reasoning mode on the efficiency-optimized Mixture-of-Experts model with a 1M-token context window, built for fast inference, high-throughput workloads, reasoning, coding, and agent workflows.
- Reasoning
- Tool Calling
- Structured Output
Added Apr 24, 2026
Model weightsPricing
Auto routing · per 1M tokens- Input
- $0.14
- Output
- $0.28
- Cache read
- $0.0028
Specifications
- Context window
- 1M
- Max output
- 384K
- Parameters
- 284B / 13B
- Total / active
- Avg output (7d)
- 1.4K tokens
- Longer than 84% of models
Benchmarks
No public benchmark scores for this model yet.
Providers
Choose or restrict providers with the API. Provider-selection prices include the selection markup. Provider selection guide · Pricing API (JSON, USD per 1K tokens)
 | $0.14 | $0.28 | $0.0028 | N/A | N/A |
| $0.14 | $0.28 | $0.0028 | N/A | N/A |
| $0.0279 | $0.472 | $0.0279 | 0.5s | 106.5 tps | |
| $0.0798$0.0718 | $0.161$0.145 | $0.0147$0.0132 | N/A | N/A | |
| $0.0882 | $0.176 | $0.0176 | 4.1s | 18 tps | |
| $0.0945 | $0.189 | $0.0189 | 1s | 32 tps | |
| $0.101 | $0.202 | $0.0206 | 1.5s | 22 tps | |
| $0.11 | $0.221 | $0.0221 | 0.7s | 72 tps | |
| $0.13 | $0.295 | N/A | N/A | N/A | |
| $0.136 | $0.242 | $0.0294 | 0.4s | 78 tps | |
| $0.147 | $0.294 | $0.0294 | N/A | N/A | |
| $0.147 | $0.294 | $0.0294 | 4.4s | 30 tps | |
| $0.147 | $0.294 | $0.0294 | 3.3s | 48 tps | |
| $0.147 | $0.294 | $0.0735 | 0.6s | 111 tps | |
| $0.157 | $0.525 | $0.0787 | 1.1s | 33 tps |
Privacy:ZDR, prompts not retainedNo training on promptsRetention and training unclearBest latency and throughput are highlighted.
Batch API
Batch rows for this model use normal routing and cost the same as standard API requests. Set provider in a row to choose a provider.