
DeepSeek V4 Pro
DeepSeek
- Context window
- 1M 1,048,576 tokens
- Input / Output
- text → text
API pricing
USD / 1M tokens
| Input | Output | Cached input |
|---|---|---|
|
|
|
Timeline
- · Weights version · evidence link
Input / output price · USD / 1M tokens
| Model | Evidence links | Input / output price |
|---|---|---|
deepseek-v4-prostandard | deepseek | $0.66 (Off-peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.); $1.32 (Peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.) / $1.98 (Off-peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.); $3.96 (Peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.) |
deepseek-v4-prostandard · text | opencode-go | $0.66 (When not (UTC days: mon, tue, wed, thu, fri and ((UTC time >= 01:00 and UTC time < 04:00) or (UTC time >= 06:00 and UTC time < 10:00)))); $1.32 (UTC days: mon, tue, wed, thu, fri and ((UTC time >= 01:00 and UTC time < 04:00) or (UTC time >= 06:00 and UTC time < 10:00))) / $1.98 (When not (UTC days: mon, tue, wed, thu, fri and ((UTC time >= 01:00 and UTC time < 04:00) or (UTC time >= 06:00 and UTC time < 10:00)))); $3.96 (UTC days: mon, tue, wed, thu, fri and ((UTC time >= 01:00 and UTC time < 04:00) or (UTC time >= 06:00 and UTC time < 10:00))) |
deepseek/deepseek-v4-prostandard | openrouter | $0.90 / $1.50 |
deepseek/deepseek-v4-prostandard | vercel-gateway | $1.32 / $3.96 |
- · Evidence links · GitHub search · DeepSeek · commits · yetone/magpie · @yetone
- · Evidence links · GitHub search · DeepSeek · commits · whistlegraph/aesthetic-computer · @whistlegraph
- · Evidence links · GitHub search · DeepSeek · commits · duketopceo/open-computer · @duketopceo
- · Evidence links · GitHub search · DeepSeek · commits · FrancisCrickInstitute/Data-Prospector · @djpbarry
- · Evidence links · GitHub search · DeepSeek · commits · Ayasesaki7/nai-relay · @Ayasesaki7
- · Evidence links · GitHub search · DeepSeek · pulls · stSoftwareAU/VibeCoder · @VibeCoderST
Excerpt
C - ->|usage limit, or no credential| D["DeepSeek deepseek-v4-pro"]
- · Evidence links · GitHub search · DeepSeek · pulls · razzant/ouroboros · @login4all
Excerpt
- Reviewer model and effort: `deepseek/deepseek-v4-pro`, high reasoning effort
- · Evidence links · GitHub search · DeepSeek · pulls · maximhq/bifrost · @jimaldon
- · Evidence links · GitHub search · DeepSeek · pulls · maximhq/bifrost · @neta79
Excerpt
- **`ui/lib/constants/config.ts`** — Model placeholder (`e.g. deepseek-chat, deepseek-v4-pro`) and key requirement entry
- · Evidence links · GitHub search · DeepSeek · pulls · maximhq/bifrost · @truffle-dev
Excerpt
returns IDs with the inference-provider segment duplicated, e.g. `huggingface/cohere/featherless-ai/deepseek-ai/DeepSeek-V4-Pro`. Requests against those IDs fail because `splitIntoModelProvider` parses the first segment (`cohere`) as the inference provider
- · Evidence links · GitHub search · DeepSeek · pulls · iv-org/invidious · @Fijxu
- · Evidence links · GitHub search · DeepSeek · pulls · iv-org/invidious · @Fijxu
- · Evidence links · GitHub search · DeepSeek · pulls · iv-org/invidious · @Fijxu
- · Evidence links · GitHub search · DeepSeek · pulls · anomalyco/models.dev · @models-dev-automation[bot]
- · Evidence links · GitHub search · DeepSeek · pulls · JSchmid6/wireguard-ops-cockpit · @JSchmid6
Excerpt
- Der in das Release eingebaute Katalog kennt `deepseek/deepseek-v4-pro` und `opencode/big-pickle`. Ohne Netz läuft opencode genauso.
- · Evidence links · DeepSeek · Research & News articles
Excerpt
Starting at 04:00 UTC on Sept 14, 2026, all deepseek-v4-pro requests will route to V4.1-Flash at V4.1-Flash rates
- · Evidence links · DeepSeek · updates
Excerpt
The official release of DeepSeek-V4-Pro will follow soon
- · Alex's posts · Alex Getman
Excerpt
⚡️ OpenAI just slashed GPT 5.6 prices Luna is now 80% cheaper, while Terra dropped 20%. The cuts also apply to subscriptions, with both models now using less of your quota. Luna is now cheaper than DeepSeek V4 Pro on input. One million tokens cost $0.20 vs $0.435.
Alex's take
Signals 2
DeepSeek V4 Pro spotted on OpenRouter
Alex's posts 1
Specifications
DeepSeek V4 Pro is a model from DeepSeek with open weights, released on 24 Apr 2026. The context window is 1,048,576 tokens. Pricing is $0.66 (Off-peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.); $1.32 (Peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.) per million input tokens and $1.98 (Off-peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.); $3.96 (Peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.) per million output tokens.
- Context window
- 1,048,576
- Input
- text
- Output
- text
- Weights
- Open
- Input · USD / 1M tokens
- $0.66 (Off-peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.); $1.32 (Peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.)
- Output · USD / 1M tokens
- $1.98 (Off-peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.); $3.96 (Peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.)
- Cached input · USD / 1M tokens
- $0.022 (Off-peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.); $0.044 (Peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.)
Questions
What is the context window of DeepSeek V4 Pro?
The context window is 1,048,576 tokens.
How much does DeepSeek V4 Pro cost?
Pricing is $0.66 (Off-peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.); $1.32 (Peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.) per million input tokens and $1.98 (Off-peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.); $3.96 (Peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.) per million output tokens. Cached input is $0.022 (Off-peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.); $0.044 (Peak: Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak, including weekends and Chinese public holidays in full.) per million.
Are the weights of DeepSeek V4 Pro publicly available?
Yes. DeepSeek publishes the weights.
When was DeepSeek V4 Pro released?
On 24 Apr 2026.