DeepSeek V4 Pro 0813 is the official release of DeepSeek's flagship V4 Pro model, superseding the April preview and rolled out simultaneously across the app, web, and API. It serves a 1M-token context window with up to 384K output tokens, supports thinking and non-thinking modes with selectable low / high / max reasoning effort, plus JSON output and tool calls, and natively speaks the Responses API with a targeted adaptation for Codex-style coding agents. The 0813 revision is a major step up in agent capability, with DeepSeek reporting especially strong gains in production environments on its published agent benchmarks — 87.9 on Terminal Bench 2.1, 83.3 on Cybergym, 74.1 on Toolathlon Verified, 62.7 on DeepSWE, 61.5 on NL2Repo, and 42.7 / 60.0 on HLE without / with tools. On the DeepSeek first-party API this revision is what the deepseek-v4-pro model id now serves; the dated identifier lists the same revision for callers that reference it by date. It is a strong pick for demanding coding agents, terminal and tool-use workloads, and complex agentic pipelines.
OpenAI-compatible — keep the SDK you already use
https://api.orcarouter.ai/v1https://api.orcarouter.aiimport os
from openai import OpenAI
client = OpenAI(
base_url="https://api.orcarouter.ai/v1",
api_key=os.environ["ORCAROUTER_API_KEY"],
)
response = client.chat.completions.create(
model="deepseek/deepseek-v4-pro-0813",
messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningreasoning_effortrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p| Input / 1M tokens · Off-peak | $0.442 |
| Output / 1M tokens · Off-peak | $0.884 |
| Cache read / 1M · Off-peak | $0.060 |
| Peak hours | 01:00–04:00, 06:00–10:00 ×2 (UTC) |
| Input / 1M tokens · ×2 | $0.884 |
| Output / 1M tokens · ×2 | $1.768 |
| Cache read / 1M · ×2 | $0.120 |
| Currency | USD |
Estimate based on list price
Estimate only — actual token counts depend on the provider's tokenizer.
What developers are saying this week
GET /api/public/models/deepseek/deepseek-v4-pro-0813Open @misc{orcarouter_deepseek_v4_pro_0813,
title = {DeepSeek V4 Pro 0813 API},
author = {DeepSeek},
year = {2026},
howpublished = {OrcaRouter},
url = {https://www.orcarouter.ai/models/deepseek/deepseek-v4-pro-0813}
}DeepSeek. (2026). DeepSeek V4 Pro 0813 API. OrcaRouter. https://www.orcarouter.ai/models/deepseek/deepseek-v4-pro-0813