DeepSeek V4 Pro (Thinking)
DeepSeek V4 Pro Thinking enables DeepSeek's chain-of-thought mode on the large-scale Mixture-of-Experts model with a 1M-token context window, built for advanced reasoning, coding, long-horizon agent workflows, knowledge, math, and software engineering tasks.
- Reasoning
- Tool Calling
- Structured Output
Added Apr 24, 2026
Model weightsPricing
Auto routing · per 1M tokens- Input
- $0.70
- Output
- $1.40
- Cache read
- $0.07
Specifications
- Context window
- 1M
- Max output
- 384K
- Parameters
- 1.6T / 49B
- Total / active
- Avg output (7d)
- 1.4K tokens
- Longer than 82% of models
Benchmarks
Benchmarks
Sourced from Vectara.
Hallucination Rate
8.6%
Factual Consistency
91.4%
Answer Rate
97.2%
Avg Summary Length
Average generated summary length
153.8
Last updated 2026-10-22 · Matched as deepseek-ai/DeepSeek-V4-Pro
Vectara LeaderboardProviders
Choose or restrict providers with the API. Provider-selection prices include the selection markup. Provider selection guide · Pricing API (JSON, USD per 1K tokens)
 | $1.10 | $2.20 | $0.11 | N/A | N/A |
| $0.70 | $1.40 | $0.07 | N/A | N/A |
| $0.735 | $1.47 | $0.0735 | 0.7s | 71 tps | |
| $1.01$0.908 | $2.02$1.82 | $0.0829$0.0747 | N/A | N/A | |
| $1.36 | $2.73 | $0.105 | 1.5s | 39 tps | |
| $1.39 | $4.16 | $0.0462 | N/A | N/A | |
| $1.50 | $2.93 | N/A | N/A | N/A | |
| $1.73 | $3.47 | $0.347 | 1.8s | 34 tps | |
| $1.76 | $3.55 | $0.136 | 1.9s | 47 tps | |
| $1.76 | $3.55 | $0.147 | N/A | N/A | |
| $1.83 | $3.65 | $0.152 | 1.3s | 41 tps | |
| $1.83 | $3.65 | $0.152 | 1.5s | 41 tps | |
| $1.84 | $3.68 | $0.147 | 1.1s | 27 tps |
Batch rows for this model use normal routing and cost the same as standard API requests. Set provider in a row to choose a provider.