DeepSeek V4 Flash
DeepSeek V4 Flash is DeepSeek's language model with a 1.3M context window and up to 1.0M output tokens, available from 16 providers, starting at $0.05 / 1M input and $0.16 / 1M output. An efficiency-optimized Mixture-of-Experts LLM from DeepSeek with 284B total and 13B activated parameters, supporting a 1M-token context window with reasoning and tool-use capabilities.
| Benchmarks | |
|---|---|
| Intelligence Index | #31 |
| Coding Index | #28 |
| GPQA | #35 |
| HLE | #38 |
| IFBench | #201 |
| Time to First Token | #112 |
| SciCode | #43 |
| LCR | #54 |
| TerminalBench Hard | #77 |
| TAU2 | #34 |
| Output TPS | #71 |
Capabilities
Input3/5
Text✓
Image✓
Audio·
Video·
PDF✓
Output1/5
Text✓
Image·
Audio·
Video·
Embedding·
Capabilities7/13
Reasoning✓
Adaptive Reasoning·
Function Calling✓
Parallel Function Calling✓
Structured Outputs✓
Native JSON Schema✓
Web Search·
URL Context·
Computer Use·
Code Execution·
File Search·
Prompt Caching✓
Assistant Prefill✓
Pricing by Provider
US Dollar ($)
Per 1M tokens
| Provider | Standard | Batch | ||||
|---|---|---|---|---|---|---|
| Input $ / 1M | Output $ / 1M | Cache Read $ / 1M | Input $ / 1M | Output $ / 1M | Cache Read $ / 1M | |
| $0.2 | $0.4 | $0.04 | $0.1 | $0.2 | N/A | |
| $0.19 | $0.51 | $0.028 | — | — | — | |
| $0.09 | $0.18 | $0.018 | — | — | — | |
| $0.44 | $1.32 | $0.014 | $0.22 | $0.66 | $0.007 | |
| $0.14 | $0.28 | $0.028 | — | — | — | |
| $0.14 | $0.28 | N/A | — | — | — | |
| $0.25 | $1.75 | N/A | — | — | — | |
| $0.14 | $0.28 | $0.028 | — | — | — | |
| $0.05 | $0.16 | $0.013 | — | — | — | |
| $0.0886 | $0.177 | $0.0177 | — | — | — | |
| $0.1 | $0.2 | N/A | — | — | — | |
| $0.2 | $0.4 | $0.04 | — | — | — | |
| $0.2 | $0.4 | $0.04 | — | — | — | |
| $0.14 | $0.28 | $0.0028 | — | — | — | |
| $0.14 | $0.28 | N/A | — | — | — | |
| $0.13 | $0.26 | $0.028 | — | — | — | |
| $0.14 | $0.28 | $0.07 | — | — | — | |
Cost Calculator
US Dollar ($)
Preset:
Cheapest Instances to Run It
Cloud GPU instances that can host DeepSeek V4 Flash, ranked by cheapest on-demand price. The model needs about 379 GB of GPU memory at FP16 precision (estimated from its parameter count), so treat the fit as guidance rather than a guarantee.
All clouds
FP16 (full precision)
US Dollar ($)
Instance | Cloud | GPU | VRAM | Price | Cheapest region | |
|---|---|---|---|---|---|---|
| g7e.24xlarge |  AWS | 4× RTX PRO Server 6000 | 384 GB | $16.57/hr | ||
| g4-standard-192 |  GCP | 4× nvidia-rtx-pro-6000 | 384 GB | $18.00/hr | ||
| p4de.24xlarge | AWS | 8× A100 | 640 GB | $27.45/hr | ||
Versions
| Version | Released | Context | Input / 1M | Output / 1M | Status |
|---|---|---|---|---|---|
| DeepSeek V4.731 Flash US | — | — | $0.424 | $1.27 | Available |
| DeepSeek V4 420 Flash | — | — | — | — | Available |
| DeepSeek V4 Flash Vision Exp | 1.0M | $0.220 | $0.660 | Available | |
| DeepSeek V4 Flash | 1.3M | $0.050 | $0.160 | Current | |
| DeepSeek V4 Flash Vision | — | — | — | — | Available |
| DeepSeek V4 Flash 731 | — | 1.0M | $0.140 | $0.280 | Available |
| DeepSeek V4 Flash Thinking | — | 200K | $0.250 | $1.75 | Available |
| DeepSeek V4 Flash US | — | — | $0.200 | $0.400 | Available |
Other Models
| Model | Tier | Released | Context | Input / 1M | Output / 1M |
|---|---|---|---|---|---|
| DeepSeek V4 Pro | Pro | 1.0M | $0.660 | $1.98 | |
| DeepSeek V4 Pro | Pro | — | — | — | — |
| DeepSeek V4 Pro | Pro | 1.0M | $0.435 | $0.870 | |
| DeepSeek V4 Pro 813 | Pro | — | 1.0M | $1.32 | $3.96 |
| DeepSeek V4 Pro US | Pro | — | — | $2.40 | $4.80 |