Not every task needs Claude Opus. Not every task can use GPT-4o mini. EnGenAI gives you granular control over which model each agent uses — per agent, per task type, per organization plan.
Most platforms pick one model for everything. That's wrong in two directions at once.
Using Opus 4.6 for every task — including updating a comment in a doc file, writing a changelog entry, or running a code review on 5 lines of CSS. You're paying premium prices for zero-value premium reasoning.
Using Haiku or GPT-4o mini for everything — including architecture decisions, security review, and designing database schemas. You get fast, cheap answers that are subtly wrong in ways that cost you weeks to fix.
The solution isn't one model. It's the right model for each task — automatically.
Assign primary and fallback models to each agent based on their role. Keith gets Opus for architecture thinking. Sage gets Gemini 2.5 Pro for documentation.
Task-type overrides apply regardless of which agent is assigned. Architecture tasks always use Opus 4.6.
Go deeper: assign by task type, not just agent. All architecture decisions → Opus 4.6. All documentation → Gemini 2.5 Pro. All code review → Sonnet 4.6.
Task-type assignments override per-agent assignments. If you set "architecture → Opus 4.6", every architecture task uses Opus regardless of which agent is assigned to it.
Task types are inferred from the task context, agent role, and PREVC phase. No manual tagging required — the routing engine classifies automatically.
Priority order: task-type override → agent assignment → org default → platform default. Every task always has a valid model target.
Plan tiers gate model access. Starter orgs use efficient models. Enterprise orgs unlock premium models and can bring their own keys.
| Model Group | Models Included | Min Plan |
|---|---|---|
| Starter | claude-haiku-4-5 gpt-4o-mini gemini-flash | Starter min tier |
| Professional | + Starter group claude-sonnet-4-6 gpt-4o gemini-2.5-pro | Pro min tier |
| Enterprise | + Pro group claude-opus-4-6 custom fine-tuned | Enterprise custom pricing |
Model groups are enforced per-organization at the platform level. A Starter organization cannot use Opus 4.6 — even if they configure it manually in their agent settings. Plan gating happens before the routing layer.
Enterprise organizations can bring their own API keys for any supported LLM provider. BYOK bypasses EnGenAI's model groups entirely — use any model available on your provider account, including custom fine-tuned models and preview APIs.
Route routine tasks to efficient models. Reserve premium models for what they're worth.
Pro and Enterprise organisations can connect their own provider API keys. Use your negotiated rates, your own rate limits, and your existing provider relationships — while keeping EnGenAI's routing, monitoring, and failover.
Store your OpenAI, Anthropic, or Google API keys. Encrypted in GCP Secret Manager — never in the database.
When you add a key, EnGenAI probes the provider to discover which models you have access to. No manual config needed.
BYO key requests bypass EnGenAI's LLM markup. You pay your provider directly at your negotiated rate.
External subscriptions — Connect your existing Claude Desktop, ChatGPT, or Google AI subscriptions. EnGenAI detects available capabilities via JSONB auto-discovery and routes accordingly.
Model assignment is the what. LLM routing is the how — health monitoring, fallback chains, and real-time rate limiting.