← Back to Docs
DeepSeek R1 32B
DeepSeek's reasoning-optimized model for complex analysis and security auditing.
| Provider | DeepSeek AI |
| Parameters | 32 billion |
| Context | 64K tokens |
| Best for | Security analysis, complex reasoning, code auditing |
| Dispatch role | security_spec_agent, verify_agent |
| Infrastructure | Dispatch Pro cloud |
| Cost | Included in Pro ($20/month, shared 5M token pool) |
Overview
DeepSeek R1 32B uses chain-of-thought reasoning to tackle problems that require deep analysis. In Dispatch, it excels at security spec generation, where it systematically evaluates attack surfaces, identifies OWASP vulnerabilities, and produces actionable security policies. It's also effective for verification tasks that need careful logical reasoning.
When Dispatch uses DeepSeek R1
- Security spec generation -- Threat modeling, attack surface analysis, security policy creation.
- Complex verification -- Multi-step reasoning about code correctness and edge cases.
- Post-build security audit -- OWASP Top 10 analysis, cross-task holistic security review.
Configuration
Available to all Dispatch Pro subscribers. Set any pipeline stage to deepseek-r1-32b in Settings > Pipeline Profiles, or let Auto Model route security-related tasks automatically.
Isolation
All cloud model runs use worktree isolation.