DeepSeek V4 Pro
DeepSeek · #3 in our ranking · Best for reasoning
DeepSeek V4 Pro is an open-weight mixture-of-experts model from DeepSeek, released Aug 2026, with 1.6T parameters, a 1M-token context window, and the MIT.
DeepSeek-V4-Pro-0813 supersedes the V4 Pro preview with much better agentic performance in production settings. It adds a DSpark speculative decoding module and is released under MIT, so commercial use and fine-tuning are straightforward.
DeepSeek V4 Pro specs
- Provider
- DeepSeek
- Architecture
- Mixture-of-experts
- Parameters
- 1.6T total
- Context
- 1M tokens | 786K words
- License
- MIT Commercial use
- Input
- Text
- Released
- Aug 2026
- Weights
- Hugging Face
Sourced from the model card on Hugging Face. Last checked 2026-10-01.
Highlights
- MIT license
- 1M-token context
- Strong reasoning and tool use
Where it fits for payment teams
- Payment reconciliation
#1 pick. Strong reasoning, a 1M-token context for whole settlement files, and an MIT license you can self-host without legal review.
- Merchant underwriting
#3 pick. Careful reasoning for borderline cases, under an MIT license.
- AML and compliance investigations
#1 pick. Strong reasoning and an MIT license, a solid default for case narratives.
Harnesses in Runtime
Run DeepSeek V4 Pro through any of these harnesses on Runtime.
Provider-reported benchmarks
| Terminal Bench 2.1 | 87.9 |
|---|---|
| Toolathlon Verified | 74.1 |
| CyberGym | 83.3 |
Benchmark scores are reported by each model provider on its Hugging Face model card. They are not independently verified by Runtime.
How to run DeepSeek V4 Pro
Download the weights
Pull deepseek-ai/DeepSeek-V4-Pro-0813 from Hugging Face. Check the MIT before you deploy.
Serve it in your cloud
Run it with vLLM or SGLang on your own GPUs, or use a managed provider that hosts the model.
Put it to work in Runtime
Point an agent or a single skill at the model through OpenCode, then compare it to your current model with evals.
DeepSeek V4 Pro FAQ
What is the context window of DeepSeek V4 Pro?
+
DeepSeek V4 Pro supports a 1M-token context window (1,048,576 tokens), roughly 786K words of English text.
How many parameters does DeepSeek V4 Pro have?
+
DeepSeek V4 Pro is a mixture-of-experts model with 1.6T total parameters.
Is DeepSeek V4 Pro free for commercial use?
+
Yes. DeepSeek V4 Pro is released under the MIT license, which is permissive and allows commercial use and fine-tuning.
What can DeepSeek V4 Pro take as input?
+
DeepSeek V4 Pro accepts text input and generates text.
Is DeepSeek V4 Pro good for payment reconciliation?
+
Yes. DeepSeek V4 Pro is our #1 pick for payment reconciliation. Strong reasoning, a 1M-token context for whole settlement files, and an MIT license you can self-host without legal review.
Is DeepSeek V4 Pro good for merchant underwriting?
+
Yes. DeepSeek V4 Pro is our #3 pick for merchant underwriting. Careful reasoning for borderline cases, under an MIT license.
Is DeepSeek V4 Pro good for AML and compliance investigations?
+
Yes. DeepSeek V4 Pro is our #1 pick for AML and compliance investigations. Strong reasoning and an MIT license, a solid default for case narratives.
Can I run DeepSeek V4 Pro in my own cloud with Runtime?
+
Yes. Download the weights from Hugging Face (deepseek-ai/DeepSeek-V4-Pro-0813) and serve them with vLLM or SGLang in your cloud, or use a managed provider that hosts the model. Runtime agents can then use DeepSeek V4 Pro through a harness like OpenCode, with your data staying in your environment.