How much does the DeepSeek V4 API cost?

DeepSeek V4 costs $0.3 per million input tokens and $0.5 per million output tokens, with cached input at $0.03 per million. A workload of 10M input and 1.5M output tokens per month costs about $3.75.

What is the context window of DeepSeek V4?

DeepSeek V4 supports a context window of 1,000,000 tokens (1M), with up to 64K output tokens per response.

Is DeepSeek V4 open source?

Yes — DeepSeek V4 is an open-weights model released under the MIT license, so it can be downloaded and self-hosted.

What are the best alternatives to DeepSeek V4?

The closest alternatives by overall capability are Qwen3-Max, GPT-5.5, Gemini 3 Pro, Gemini 2.5 Pro. See the comparison pages for detailed head-to-head breakdowns.

DeepSeek V4

Name: DeepSeek V4
Price: 0.3 USD
Author: DeepSeek

DeepSeek·Mar 2026reasoningopen weights · MIT

DeepSeek V4 did to 2026 what its R1 did to 2025: proved that open-weights models can sit at the frontier. It scores around 81% on SWE-bench Verified — within a few points of Claude Opus 4.8 — with a 1M-token context, hybrid reasoning modes, and API pricing an order of magnitude below US flagships. Under MIT license, it can be self-hosted, fine-tuned and embedded commercially without restriction, making it the default open frontier model for cost-sensitive and sovereignty-sensitive deployments alike.

Benchmark results

SWE-bench VerifiedCoding · #7 of 33~81%

Where it shines

Frontier-class coding within a few points of closed flagships
10–50x cheaper than comparable proprietary APIs
MIT license — unrestricted commercial use and fine-tuning
1M-token context window

Alternatives to DeepSeek V4

Qwen3-Max

Alibaba's trillion-parameter API flagship — frontier-adjacent quality with strong agentic tool use at mid-tier prices.

GPT-5.5

OpenAI's flagship reasoning model with a 1M-token context window, built for hard coding, science and long-horizon agentic work.

Gemini 3 Pro

Google's November 2025 frontier breakout — 91.9% GPQA Diamond and 37.5% HLE made it the reasoning leader of its generation.

Gemini 2.5 Pro

Google's 2025 workhorse flagship — first mainstream thinking model with a 1M context, still widely deployed.

Frequently asked questions

How much does the DeepSeek V4 API cost?: DeepSeek V4 costs $0.3 per million input tokens and $0.5 per million output tokens, with cached input at $0.03 per million. A workload of 10M input and 1.5M output tokens per month costs about $3.75.
What is the context window of DeepSeek V4?: DeepSeek V4 supports a context window of 1,000,000 tokens (1M), with up to 64K output tokens per response.
Is DeepSeek V4 open source?: Yes — DeepSeek V4 is an open-weights model released under the MIT license, so it can be downloaded and self-hosted.
What are the best alternatives to DeepSeek V4?: The closest alternatives by overall capability are Qwen3-Max, GPT-5.5, Gemini 3 Pro, Gemini 2.5 Pro. See the comparison pages for detailed head-to-head breakdowns.