Auto Router (Beta)
Auto Router (Beta) is a task-aware router from OpenRouter. It classifies each request, then routes it the [most popular model](/rankings#task-spend) for that task based on aggregate spend, filtered by your...
Anyone in the Project can @-mention Auto Router (Beta) with the team's shared context - pooled credits, one chat, one memory.
Starter is free forever - 1 Project, 100 credits/month, 1 MCP. No card.
Verdict
Best for
- Prototyping with minimal model research
- Cost-sensitive batch processing tasks
- Exploratory data analysis workflows
- Low-stakes content generation at scale
Strengths
Auto Router eliminates model selection paralysis by routing requests to whichever backend offers the best speed-cost-availability mix at query time. The 2M token context window suggests it can handle long documents when routed to appropriate models. Multi-modal support (text, image, audio, video, file) means you can throw varied input types at it without pre-filtering. For teams running thousands of requests where individual output variance is acceptable, this reduces both decision fatigue and average spend.
Trade-offs
You lose reproducibility — the same prompt may hit different models on successive runs, making A/B testing or debugging nearly impossible. Pricing is opaque (listed as unknown), so cost forecasting requires post-hoc analysis. Without public benchmarks, you're flying blind on capability floors; a critical prompt might land on a weaker model during peak load. Teams needing consistent voice, format adherence, or audit trails will find this router frustrating. The beta label signals expect breaking changes and limited support.
Specifications
- Provider
- openrouter
- Category
- image
- Context length
- 2,000,000 tokens
- Max output
- —
- Modalities
- text, image, audio, file, video
- License
- proprietary
- Released
- 2026-07-17
Pricing
- Input
- —
- Output
- —
- Model ID
openrouter/auto-beta
Per-token prices show what the model costs upstream. On Switchy your team draws from one shared org credit pool - one plan, one balance for everyone.
Providers
Performance
Benchmarks
Works well with
Top MCPs
Compatibility data comes from first-party telemetry; once we have enough co-usage signal, top MCPs for this model will appear here.
How Switchy teams use it
Starter prompts
Summarize Research PDFs
Read this PDF and provide a 200-word summary covering the research question, methodology, and main findings. Focus on actionable insights.Open in a Project →
Classify Support Tickets
Classify this support ticket into one of these categories: billing, technical, feature request, or other. Provide the category name and a one-sentence justification.Open in a Project →
Generate Product Descriptions
Write a 50-word product description for this item emphasizing benefits over features. Use an enthusiastic but professional tone.Open in a Project →
Extract Data from Screenshots
Extract all visible text, numbers, and labels from this screenshot into a JSON object. Use descriptive keys for each field you identify.Open in a Project →
Brainstorm Campaign Ideas
Suggest five marketing campaign ideas for this product targeting small business owners. For each idea, include a headline and two-sentence pitch.Open in a Project →
Compare with
More image models
- Google: Nano Banana 2 (Gemini 3.1 Flash Image)google
- Google: Nano Banana 2 (Gemini 3.1 Flash Image Preview)google
- Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)google
- Google: Nano Banana (Gemini 2.5 Flash Image)google
- Google: Nano Banana Pro (Gemini 3 Pro Image)google
- Google: Nano Banana Pro (Gemini 3 Pro Image Preview)google
- OpenAI: GPT-5.4 Image 2openai
- OpenAI: GPT-5 Imageopenai
- OpenAI: GPT-5 Image Miniopenai
- Auto Routeropenrouter