Canada's sovereign frontier AI lab building the Raven models and DCode
OpenAI-compatible inference API with 92 models at discounted prices
Stage
Startup
Not given
Founded
2026
Not given
Based in
CA
Not given
Pricing model
Freemium
Paid
Pricing
Raven Free (free forever with automatically replenishing limits), Pro ($20/month or $200/year, Founder $160/year), Ultra ($100/month or $1,000/year, Founder $800/year). DCode Coding Plan $29/month (Founder $230/year). API: Raven Flash $0.50 per 1M input tokens, $2.50 per 1M output tokens; Raven Max
Pay as you go prepaid credit starting at $5 minimum top-up. Free $0.05 credit with no card required. Claude Opus 5.5: $1.00 input / $5.00 output per 1M tokens. Claude Sonnet 5.5: $0.50 input / $2.50 output per 1M tokens. Claude Haiku 5.5: $0.025 input / $0.125 output per 1M tokens. GPT-6.1 Sol: $2.0
Key features
Native multimodal vision for diagrams, schematics, and error logs
Autonomous tool execution with structured receipts
Inspectable reasoning streams in real time
1M token context length
OpenAI and Anthropic SDK compatibility
Automatic prompt caching with up to 90% input cost savings
Model Context Protocol (MCP) support
Canadian sovereign compute and PIPEDA compliance
Real-time streaming and sub-second latency
Prepaid credits that never expire
92 models from multiple providers
OpenAI-compatible endpoint
Streaming, tool calling, structured output
CLI for remote machine management
Scoped API keys for machine access
Free trial with $0.05 credit
JSON output and exit codes
Local MCP tools for agents
Private cloud machines with agent integration
Billing per minute with pause functionality
What makes it different
Independent frontier AI built and served from Canada with statutory PIPEDA data privacy
Inspectable reasoning streams for transparent, auditable AI decisions
#1 ranked on Terminal Bench, Agent's Last Exam, and AutomationBench benchmarks
Bootstrapped from first principles without venture capital
Unified continuous architecture across multiple frontier models
75% below published Claude list prices while maintaining model fidelity
Single endpoint and key for 92 frontier and open models
Drop-in OpenAI replacement requiring only base_url change
AI agents with preinstalled tools and no API key configuration needed
Infrastructure company with dedicated H100/H200 capacity roadmap