DeepSeek V4 Flash Vision Exp
Summary
DeepSeek V4 Flash Vision Exp is an open-weight language model from DeepSeek, released on 21 Aug 2026. It accepts text and images and generates text, with a 1M-token context window and up to 1M output tokens per response. As of 18 Nov 2026, the lowest listed API price is $0.22 per million input tokens and $0.65 per million output tokens (Vercel AI Gateway), across 5 providers we track.
Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work
Last verified 18 Nov 2026. Every figure on this page links to its source.
Specifications
| Developer | DeepSeek |
|---|---|
| Release date | 21 Aug 2026 [models.dev] |
| Status | Generally available [models.dev] |
| Weights | Open weights [Hugging Face] |
| License | mit [Hugging Face] |
| Parameters | 305B [Hugging Face] |
| Context window | 1M (1,048,576 tokens) [models.dev] |
| Max output | 1M tokens [models.dev] |
| Input | Text and images [models.dev] |
| Output | Text [models.dev] |
| Reasoning mode | Yes [models.dev] |
| Tool calling | Yes [models.dev] |
| Structured output | Yes [models.dev] |
| Hugging Face | deepseek-ai/DeepSeek-V4-Flash-Vision-Exp [Hugging Face] |
DeepSeek V4 Flash Vision Exp API pricing by provider
USD per million tokens. Sorted by input price.
| Provider | Input | Output | Cached input | Context | Source |
|---|---|---|---|---|---|
| Vercel AI Gateway deepseek/deepseek-v4-flash-vision-exp | $0.22 | $0.65 | $0.0068 | 1M | models.dev, 18 Nov 2026 |
| DeepInfra deepseek-ai/DeepSeek-V4-Flash-Vision-Exp | $0.44 | $1.32 | $0.01 | 1M | models.dev, 18 Nov 2026 |
| Hugging Face Inference Providers deepseek-ai/DeepSeek-V4-Flash-Vision-Exp | $0.44 | $1.32 | – | 1M | models.dev, 18 Nov 2026 |
| SiliconFlow deepseek-ai/DeepSeek-V4-Flash-Vision-Exp | $0.44 | $1.32 | $0.03 | 1M | models.dev, 18 Nov 2026 |
| OpenRouter deepseek/deepseek-v4-flash-vision-exp | $0.44 | $1.32 | $0.01 | 1M | OpenRouter, 18 Nov 2026 |
Running DeepSeek V4 Flash Vision Exp locally
Estimated memory, from the parameter count and the architecture in config.json on Hugging Face. Includes about 0.8 GB of runtime overhead.
| Quantization | Weights | Total, 8K context | Total, 32K context |
|---|---|---|---|
| Q4_K_M (4-bit) | 185 GB | 186 GB | 188 GB |
| Q8_0 (8-bit) | 324 GB | 325 GB | 327 GB |
| FP16 / BF16 (16-bit) | 609 GB | 611 GB | 613 GB |
At 4-bit with an 8K conversation, DeepSeek V4 Flash Vision Exp needs about 186 GB, more than any single consumer graphics card; it needs a unified-memory machine, a multi-GPU setup or a split between graphics card and system RAM. Check your own hardware, or see Best computer for local AI on our sister site AIPCs.
Frequently asked questions
What is the context window of DeepSeek V4 Flash Vision Exp?
DeepSeek V4 Flash Vision Exp has a context window of 1,048,576 tokens (1M) and can generate up to 1,048,576 tokens in a single response, according to models.dev.
How much does the DeepSeek V4 Flash Vision Exp API cost?
DeepSeek V4 Flash Vision Exp is listed from $0.22 per million input tokens and $0.65 per million output tokens (Vercel AI Gateway). Prices last verified 18 Nov 2026.
Is DeepSeek V4 Flash Vision Exp open source?
DeepSeek V4 Flash Vision Exp is an open-weight model: its weights are published on Hugging Face as deepseek-ai/DeepSeek-V4-Flash-Vision-Exp under the mit license. Check the license terms before commercial use.
When was DeepSeek V4 Flash Vision Exp released?
DeepSeek released DeepSeek V4 Flash Vision Exp on 21 Aug 2026, according to models.dev.
Which providers offer DeepSeek V4 Flash Vision Exp?
We track DeepSeek V4 Flash Vision Exp on 5 providers: Vercel AI Gateway, DeepInfra, Hugging Face Inference Providers, SiliconFlow, OpenRouter.
How much VRAM does DeepSeek V4 Flash Vision Exp need?
About 186 GB at 4-bit (Q4_K_M) with an 8K-token conversation, and about 325 GB at 8-bit. That is more than any single consumer graphics card holds. These are estimates from the model's published parameter count and architecture.
Similar models from other labs
Comparable on output type, weights, release date, price and context window. Prices are the lowest listed, USD per million tokens.
| Model | Lab | Released | Context | Input $/1M | Output $/1M | Weights |
|---|---|---|---|---|---|---|
| GLM-5.3-Flash | Zhipu AI (Z.ai) | 26 Aug 2026 | 1M | $0.15 | $0.50 | Open |
| GPT-6 Luna Pro | OpenAI | 22 Sept 2026 | 1.1M | $0.10 | $0.50 | – |
| MiMo-V2.6-Pro | Xiaomi (MiMo) | 22 Sept 2026 | 1M | $0.43 | $0.87 | Open |
| NVIDIA Nemotron 3.5 Super VL | NVIDIA | 15 Oct 2026 | 1M | $0.15 | $0.50 | Open |
| GLM-5.3 | Zhipu AI (Z.ai) | 14 Aug 2026 | 1M | $0.07 | $1.40 | Open |
| Beam 1.0 | Reflection | 5 Oct 2026 | 1M | $0.25 | $1 | Open |
Change history
- Max output changed from 384K to 1M
- Context window changed from 1M to 1M
- Context window changed from 1M to 1M
- Max output changed from 1M to 384K
- Max output changed from 384K to 1M
- Max output changed from 1M to 384K
- Max output changed from 384K to 1M
- Context window changed from 1M to 1M
- Max output changed from 944K to 384K
- Context window changed from 1M to 1M
- OpenRouter price changed: input $0.22 → $0.44, output $0.65 → $1.32 per 1M tokens
- Max output changed from 384K to 944K
- Context window changed from 1M to 1M
- Context window changed from 1M to 1M
- Max output changed from 262K to 384K
- Max output changed from 384K to 262K
- Context window changed from 1M to 1M
- Context window changed from 1M to 1M
- Max output changed from 262K to 384K
- Max output changed from 384K to 262K
- Context window changed from 1M to 1M
- Context window changed from 1M to 1M
- Max output changed from 262K to 384K
- Max output changed from 1M to 262K
- Max output changed from 384K to 1M
Sources
- https://openrouter.ai/deepseek/deepseek-v4-flash-vision-exp
- https://github.com/sst/models.dev/blob/dev/providers/vercel/models/deepseek/deepseek-v4-flash-vision-exp.toml
- https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp
- https://github.com/sst/models.dev/blob/dev/providers/deepinfra/models/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp.toml
- https://github.com/sst/models.dev/blob/dev/providers/huggingface/models/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp.toml
- https://github.com/sst/models.dev/blob/dev/providers/siliconflow/models/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp.toml