> ## Documentation Index > Fetch the complete documentation index at: https://www.stratus.run/docs/llms.txt > Use this file to discover all available pages before exploring further. # What is Stratus? Stratus X1 is the first predictive action model purpose-built for AI agents — turning unreliable, expensive agents into systems that plan before they act and succeed instead of guess. Stratus X1 is a **predictive action model** that sits between your LLM and the environment. It understands where an agent is, simulates what happens next, and sequences actions toward a goal — before a single real action executes. ## The Problem with Current Agents LLM-based agents achieve only **10–20% success rates** on real-world benchmarks like WebArena. Human performance on the same tasks: **78%**. The gap is structural, not a prompting problem. Current agents fail for four compounding reasons: Agents can't represent where they are in a task. Every step starts from scratch with raw, noisy context. Without a world model, agents can't foresee what an action will do before committing to it. Multi-step planning is impossible when there's no mechanism for reasoning over state sequences. Processing raw tokens at 15,000+ per task is slow, expensive, and fundamentally noisy. The result: low success, high cost, high latency, and brittle plans with no recovery path. ## The Stratus Solution Stratus doesn't replace your LLM. It makes your LLM dramatically more effective by handling the parts LLMs are fundamentally bad at — state representation, consequence modeling, and action sequencing. Stratus encodes any observation — a webpage, a UI state, a tool response — into a rich semantic representation. The noise disappears. The meaning stays. The world model predicts what the environment looks like after each candidate action, in representation space, before anything executes. Your agent sees the future before committing. The planning layer sequences actions toward the goal using the world model's predictions. It returns a ranked plan with a confidence score at each step, so your LLM reasons over structure — not noise. ## Three Components, One Coherent System Compresses any environment description into a rich representation. The richer your state, the sharper the plan. Simulates what the environment looks like after each action — entirely in representation space, before anything executes. Sequences actions toward the goal using the world model's predictions. Returns a ranked plan with confidence at each step. ## Without Stratus vs. With Stratus Raw observations flood the LLM with 15,000+ tokens of noisy context. The model guesses at each step, with no ability to foresee consequences or recover from mistakes. Success rates hover at 10–20%. Every failure is expensive. Stratus extracts meaning from the environment, simulates candidate actions, and hands the LLM a structured plan. Token count drops by over 60%. Success rates double. Failures are predictable and recoverable. **The measured difference:** 68% fewer tokens. 2–3x faster. 2x+ higher task success rate. And failure modes you can actually reason about. ## Why This Is Different Stratus operates in **representation space** — not token space. This is a fundamental architectural distinction from every other approach: RAG retrieves documents. Stratus learns state transitions. Retrieval doesn't tell you what happens next — a world model does. Better prompts reorder text. Stratus predicts outcomes in embedding space. No prompt can teach an LLM to simulate consequences. Fine-tuning adjusts token distributions. Stratus models state — what exists, what changes, what's next. These are categorically different problems. ## Performance Stratus delivers measurable results from day one — not theoretical improvements on held-out benchmarks. Over **20x reduction** in tokens consumed per task. What took 15,000 tokens now takes under 750. Better than **75% detection rate** on hallucinated actions and fabricated state — caught before they execute. **Under 10ms** per prediction. Stratus adds no meaningful latency to your agent loop. **1,000+ predictions per second** — scales with your workload, not against it. ## Where Stratus Excels Booking flows, form completion, data extraction — tasks where multi-step state tracking is everything. Research tasks that require chaining searches, synthesizing results, and maintaining goal context across many steps. Workflow automation, software testing, data entry — high-volume tasks where reliability directly translates to cost. Any environment with predictable state transitions — APIs, UIs, robotic action spaces — where consequence modeling compounds. ## Model Tiers | Model | Best For | Prediction Latency | | ------- | ------------------------------------------------- | ------------------ | | `small` | Highest throughput, latency-sensitive loops | Under 10ms | | `base` | Balanced performance — recommended starting point | Under 25ms | | `large` | Extended context tasks | Under 50ms | | `xl` | Maximum context and precision | Under 100ms | ## What Comes Next Text-based meaning model for software agents. Production-ready, OpenAI API-compatible, integrates with any agent framework in minutes. Multimodal world model: text, vision, and telemetry. Extends Stratus into robotics and physical systems. Full world-model layer for multi-agent systems. Stratus becomes the meaning layer for all autonomous infrastructure. Ready to see it in action? The [Quickstart](/docs/quickstart) gets you running in under five minutes. Or go deeper with the [API Reference](/docs/api-reference/chat-completions).