All models
DeepSeek
DeepSeek V4 Pro 0813 frontierruns locally
deepseek/deepseek-v4-pro-0813
DeepSeek V4 Pro 0813 is a large-scale mixture-of-experts model from DeepSeek. This is the GA release of DeepSeek V4 Pro.
Input / 1M
$0.12
USD per million tokens
Output / 1M
$4.60
USD per million tokens
Est. cost / token
$0.000001236
$1.24 per 1M, 3:1 mix
Context window
1.05M
max output 944K
Major features
Open weightsCachingReasoningJSON schemaTools
Task suitability (heuristic 1–5, not a benchmark)
- coding5
- reasoning5
- math5
- summarization5
- extraction5
- classification5
- chat5
- creative5
- translation5
- vision1
- long context5
- agentic5
Run it locally
Will it run on my machine? →| Quantization | Quality | Memory needed (8K ctx) |
|---|---|---|
| Q2_K | poor | 736 GB |
| Q3_K_M | fair | 858 GB |
| Q4_K_Mrecommended | good | 1070 GB |
| Q5_K_M | very good | 1243 GB |
| Q6_K | excellent | 1432 GB |
| Q8_0 | near-lossless | 1852 GB |
| F16 | lossless | 3477 GB |
Estimates.
Price history
| Recorded | Input | Output | Cached | Source |
|---|---|---|---|---|
| 2026-11-05 03:20 | $0.12 | $4.60 | $0.11 | openrouter |
A row is added whenever a sync detects a price change. Vendor pricing source
Pricing (USD per 1M tokens)
- Input
- $0.12
- Output
- $4.60
- Cached input
- $0.11
- Blended (3:1)
- $1.24
Details
- Company
- DeepSeek
- Version
- 4
- Family
- DeepSeek
- Released
- 2026-08-12
- Parameters
- 1.65T
- Max output
- 944K
- Open weights
- Yes
- Inputs
- text
- Last synced
- 21 min ago