Per Section 3.5 of the Terms of Service, this page details the AI model portfolio available within Entropy IDE, along with context window limits and consumption rates applicable when exceeding included monthly quotas.
| Model | Provider | Context Limit | Input Rate | Output Rate | Ideal Use Case |
|---|---|---|---|---|---|
Claude 3.7 SonnetPOPULAR | Anthropic | 200k tokens | R55.00 / 1M tokens | R275.00 / 1M tokens | Complex agentic multi-file refactoring & hybrid thinking |
Claude 3.5 Haiku | Anthropic | 200k tokens | R15.00 / 1M tokens | R75.00 / 1M tokens | Lightning-fast inline suggestions & doc generation |
GPT-4o | OpenAI | 128k tokens | R45.00 / 1M tokens | R180.00 / 1M tokens | Full-stack code generation, Python & TypeScript |
o3-mini | OpenAI | 200k tokens | R20.00 / 1M tokens | R80.00 / 1M tokens | Algorithmic puzzles, math, and compiler design |
Gemini 2.5 FlashPOPULAR | 1,000k tokens (1M) | R2.80 / 1M tokens | R11.00 / 1M tokens | Massive codebase semantic search & monorepo indexing | |
Gemini 2.5 Pro | 2,000k tokens (2M) | R23.00 / 1M tokens | R92.00 / 1M tokens | Deep monorepo refactoring & whole-project architecture | |
DeepSeek R1 | DeepSeek | 64k tokens | R10.00 / 1M tokens | R40.00 / 1M tokens | Open reasoning chains and zero-shot algorithm design |
Apolemia Real-time InfillPOPULAR | Knolink (Pty) Ltd | 16k tokens | Included in Pro / Free (BYOK) | Included in Pro / Free (BYOK) | Instant sub-10ms ghost-text tab autocomplete |
Bring Your Own Key (BYOK) Freedom
Entropy never locks you into proprietary inference margins. In both the Desktop IDE and Web Editor, you can paste your personal Anthropic, OpenAI, or Google AI Studio API key in Settings → AI & Inference to run all generations at direct zero-margin provider rates.
API keys are encrypted locally with machine-level OS credentials.