DeepSeek: DeepSeek V4 Pro 0423
The DeepSeek V4 Pro 0423 is a language model developed by deepseek that handles text input modalities up to 1 MB in context length. It supports tools and reasoning capabilities but does not output structured results. This model may be worth considering for those with specific requirements who value its performance on tasks covered by the blended benchmark score of 78.0 across two independent benchmarks. However, at $3.2 per million output tokens, it is one of the more expensive options available; those prioritizing cost-effectiveness should carefully weigh this price against their needs.
Benchmark results
Independent, published benchmarks. Blended score 84.4 across 4 benchmarks, last refreshed 2026-11-17. How scoring works →
| Benchmark | Measures | Score |
|---|---|---|
| AI Index | broad capability composite | 73.1 |
| AI Index Coding | software engineering tasks | 97.9 |
| AI Index Agentic | multi-step tool-using tasks | 60.0 |
| EQ-Bench | emotional understanding in dialogue | 77.6 |
- Model ID
- deepseek/deepseek-v4-pro
- Vendor
- deepseek
- Released
- April 2026
- Tokenizer
- DeepSeek
- Input Modalities
- text
- Output Modalities
- text
- Max Output
- 393,216 tokens
- Tool Calling
- ✓ supported
- Structured Output
- ✓ supported
- Reasoning Mode
- ✓ supported
- Vision
- text only
- Audio
- no
- Moderated
- no
What it costs in practice
Computed from the current $0.90/M input and $1.50/M output rates. Run your own numbers →
| Job | Tokens | Cost |
|---|---|---|
| Summarize a 50-page report | 30k in / 1.5k out | $0.03 |
| Classify 1,000 customer emails | 500k in / 50k out | $0.53 |
| A month of a busy support chatbot | 5M in / 2M out | $7.50 |
Price & spec history
Tracked daily by PicksByModel since 2026-07-17.
| Date | Input /M | Output /M | Context |
|---|---|---|---|
| 2026-11-14 | $0.90 | $1.50 | 1,048,576 |
| 2026-11-09 | $0.96 | $1.91 | 1,048,576 |
| 2026-11-08 | $0.43 | $0.87 | 1,048,576 |
| 2026-11-07 | $0.89 | $1.78 | 1,048,576 |
| 2026-11-06 | $0.96 | $1.91 | 1,048,576 |
| 2026-10-31 | $1.60 | $3.20 | 1,048,576 |
| 2026-10-30 | $0.82 | $1.63 | 1,048,576 |
| 2026-10-26 | $0.96 | $1.91 | 1,048,576 |
| 2026-10-25 | $0.21 | $0.42 | 1,048,576 |
| 2026-10-24 | $0.59 | $1.18 | 1,048,576 |
| 2026-10-19 | $0.96 | $1.91 | 1,048,576 |
| 2026-10-17 | $0.21 | $0.42 | 1,048,576 |
| 2026-10-12 | $0.96 | $1.91 | 1,048,576 |
| 2026-10-11 | $0.21 | $0.42 | 1,048,576 |
| 2026-10-10 | $0.95 | $1.90 | 1,048,576 |
| 2026-10-08 | $0.96 | $1.91 | 1,048,576 |
| 2026-10-02 | $0.21 | $0.42 | 1,048,576 |
| 2026-10-01 | $0.26 | $0.51 | 1,048,576 |
| 2026-09-28 | $0.96 | $1.91 | 1,048,576 |
| 2026-09-27 | $0.35 | $0.70 | 1,048,576 |
| 2026-09-26 | $0.42 | $0.84 | 1,048,576 |
| 2026-09-25 | $0.81 | $1.62 | 1,048,576 |
| 2026-09-24 | $0.94 | $1.88 | 1,048,576 |
| 2026-09-21 | $0.96 | $1.91 | 1,048,576 |
| 2026-09-20 | $0.42 | $0.84 | 1,048,576 |
| 2026-09-19 | $0.62 | $1.25 | 1,048,576 |
| 2026-09-18 | $0.94 | $1.89 | 1,048,576 |
| 2026-09-13 | $1.60 | $3.20 | 1,048,576 |
| 2026-09-12 | $0.83 | $1.66 | 1,048,576 |
| 2026-09-08 | $0.96 | $1.91 | 1,048,576 |
| 2026-09-07 | $1.04 | $2.08 | 1,048,576 |
| 2026-09-06 | $0.76 | $1.52 | 1,048,576 |
| 2026-09-05 | $0.96 | $1.92 | 1,048,576 |
| 2026-09-02 | $1.04 | $2.08 | 1,048,576 |
| 2026-09-01 | $1.60 | $3.20 | 1,048,576 |
| 2026-08-29 | $0.70 | $1.40 | 1,048,576 |
| 2026-08-27 | $0.87 | $1.74 | 1,048,576 |
| 2026-08-25 | $0.79 | $1.58 | 1,048,576 |
| 2026-08-24 | $0.53 | $1.05 | 1,048,576 |
| 2026-08-23 | $0.41 | $0.83 | 1,048,576 |
| 2026-08-22 | $0.50 | $1.00 | 1,048,576 |
| 2026-08-21 | $1.60 | $3.20 | 1,048,576 |
| 2026-08-20 | $1.44 | $2.88 | 1,048,576 |
| 2026-08-17 | $0.66 | $1.98 | 1,048,576 |
| 2026-08-13 | $1.17 | $2.34 | 1,048,576 |
| 2026-08-12 | $0.63 | $1.26 | 1,048,576 |
| 2026-08-11 | $0.69 | $1.37 | 1,048,576 |
| 2026-07-17 | $0.43 | $0.87 | 1,048,576 |
Strong choice for
Category rankings
Where DeepSeek: DeepSeek V4 Pro 0423 places across the 62 categories it ranks in. How we rank →
| # | Category | Score |
|---|---|---|
| #2 | Code CompletionCode · of 25 ranked | 134 |
| #5 | Transcript CleanupWriting · of 25 ranked | 145 |
| #5 | Meeting NotesBusiness · of 25 ranked | 157 |
| #6 | Short-Form SummarizationWriting · of 25 ranked | 131 |
| #6 | Email DraftingWriting · of 25 ranked | 127 |
| #6 | Language LearningEducation · of 25 ranked | 127 |
| #6 | Chat CompanionPersonal · of 25 ranked | 131 |
| #8 | Code DocumentationCode · of 25 ranked | 146 |
| #8 | Trivia & General KnowledgePersonal · of 25 ranked | 121 |
| #10 | CSV / Spreadsheet CleanupData · of 25 ranked | 156 |
| #10 | Long-Document SummarizationWriting · of 25 ranked | 165 |
| #10 | Blog Post WritingWriting · of 25 ranked | 142 |
| #10 | Character RoleplayPersonal · of 25 ranked | 154 |
| #10 | Trip PlanningPersonal · of 25 ranked | 142 |
| #11 | ETL ScriptingData · of 25 ranked | 157 |
| #11 | Medical Note SummarizationProfessional · of 25 ranked | 157 |
| #11 | History TutoringEducation · of 25 ranked | 148 |
| #11 | Long-Context Q&AAgents · of 25 ranked | 160 |
| #12 | Code ReviewCode · of 25 ranked | 170 |
| #12 | Code RefactoringCode · of 25 ranked | 169 |
| #12 | Creative WritingWriting · of 25 ranked | 161 |
| #12 | Screenwriting & DialogWriting · of 25 ranked | 161 |
| #12 | Academic WritingWriting · of 25 ranked | 161 |
| #12 | Legal DraftingProfessional · of 25 ranked | 161 |
| #12 | Legal ResearchProfessional · of 25 ranked | 173 |
| #12 | Contract ReviewProfessional · of 25 ranked | 182 |
| #12 | Scientific ResearchProfessional · of 25 ranked | 173 |
| #12 | Essay GradingEducation · of 25 ranked | 170 |
| #12 | RFP ResponseBusiness · of 25 ranked | 179 |
| #12 | OKRs & Strategic PlanningBusiness · of 25 ranked | 161 |
| #12 | RAG PipelinesAgents · of 25 ranked | 169 |
| #12 | Fiction CollaboratorPersonal · of 25 ranked | 173 |
| #12 | Math ProofsResearch · of 25 ranked | 161 |
| #12 | Literature ReviewResearch · of 25 ranked | 173 |
| #13 | Physics TutoringEducation · of 25 ranked | 149 |
| #13 | Sales / Cold EmailBusiness · of 25 ranked | 123 |
| #13 | Job Application DraftingBusiness · of 25 ranked | 123 |
| #13 | Journaling HelperPersonal · of 25 ranked | 123 |
| #13 | Recipe GenerationPersonal · of 25 ranked | 123 |
| #14 | SQL GenerationCode · of 25 ranked | 171 |
| #14 | Unit Test GenerationCode · of 25 ranked | 155 |
| #14 | Regex WritingCode · of 25 ranked | 136 |
| #14 | CI/CD PipelinesCode · of 25 ranked | 146 |
| #14 | JSON ExtractionData · of 25 ranked | 146 |
| #14 | Language TranslationWriting · of 25 ranked | 136 |
| #14 | Tutoring for KidsPersonal · of 25 ranked | 136 |
| #15 | Customer SupportBusiness · of 25 ranked | 132 |
| #16 | Bug FixingCode · of 25 ranked | 182 |
| #16 | Scientific CodingResearch · of 25 ranked | 182 |
| #17 | Dataset AnnotationResearch · of 25 ranked | 143 |
| #18 | Marketing CopyWriting · of 25 ranked | 130 |
| #18 | Standardized Test PrepEducation · of 25 ranked | 130 |
| #18 | Agent WorkflowsAgents · of 25 ranked | 192 |
| #18 | Coding AgentsAgents · of 25 ranked | 192 |
| #20 | Bulk Data LabelingData · of 25 ranked | 133 |
| #22 | Resume WritingBusiness · of 25 ranked | 117 |
| #22 | SOP / Process DocsBusiness · of 25 ranked | 136 |
| #23 | Fitness CoachingPersonal · of 25 ranked | 134 |
| #24 | Data AnalysisData · of 25 ranked | 168 |
| #24 | Financial AnalysisProfessional · of 25 ranked | 157 |
| #24 | Math TutoringEducation · of 25 ranked | 147 |
| #24 | Experiment DesignResearch · of 25 ranked | 156 |
Similar models
DeepSeek: DeepSeek V4 Pro 0813
DeepSeek: DeepSeek V4 Flash 0731
DeepSeek: DeepSeek V4 Flash 0731 (batch)
DeepSeek: DeepSeek V4 Flash 0423
DeepSeek: DeepSeek V4.1 Flash 1017
DeepSeek: DeepSeek V4.1 Flash 1017 (batch)
Quick answers
- How much does DeepSeek: DeepSeek V4 Pro 0423 cost?
- $0.90 per million input tokens and $1.50 per million output tokens.
- What is DeepSeek: DeepSeek V4 Pro 0423's context window?
- 1,048,576 tokens, roughly 1,572 pages of text in a single request.
- Does DeepSeek: DeepSeek V4 Pro 0423 support tool calling?
- It supports tool calling, structured output, a reasoning mode.
- Can DeepSeek: DeepSeek V4 Pro 0423 process images?
- No, it is text-only on the input side.