México
Colombia
BrasilRouting playground
Describe your monthly AI traffic and see which model each request would go to, and what it would cost.
Model prices are each provider's current list prices. Quality scores and latency are illustrative samples, not benchmarks. Nothing is sent to an AI provider.
Your workload
How far below your current model's quality score the router may go.
Today: $746/mo. With Clara Router: $501/mo. Saved: 32.8%.
- Today
- $746/mo
Everything on GPT-6.1 Sol
- With Clara Router
- $501/mo
- Saved
- 32.8%
$245/mo
Routing plan
Each task goes to the model that best fits your policy without falling below the quality bar.
| Task | Traffic | Today | Routed to | Per request | Per month |
|---|---|---|---|---|---|
| 60% | $0.0026 | DeepSeek V4 Pro | $0.00166 | $100 | |
| 30% | $0.0104 | No change | $0.0104 | $312 | |
| 10% | $0.0278 | Grok 4.7 | $0.0218 | $218 |
Models for this task: Ticket triage
1,200 tokens in · 20 out
Sample request
Clasifica la urgencia de este ticket: Mi tarjeta empresarial fue rechazada y necesito pagar a un proveedor hoy.
Quality bar 97: the score of GPT-6.1 Sol on this task, minus your tolerance.
| Model | Price per 1M in / out | Per request | Latency | Quality | Decision |
|---|---|---|---|---|---|
| Claude Haiku 5.5 | $0.10 / $0.50 | $0.00013 | 0.4 s | 96 | Below bar |
| GPT-6 Luna | $0.10 / $0.50 | $0.00013 | 0.4 s | 95 | Below bar |
| Qwen3.8-Flash | $0.15 / $0.47 | $0.000189 | 0.4 s | 95 | Below bar |
| Mistral Small 4 | $0.15 / $0.60 | $0.000192 | 0.3 s | 94 | Below bar |
| Muse Spark 1.3 | $1.25 / $4.25 | $0.00159 | 0.6 s | 96 | Below bar |
| DeepSeek V4 Pro | $1.32 / $3.96 | $0.00166 | 1.2 s | 97 | Selected |
| GPT-6.1 SolToday | $2 / $10 | $0.0026 | 0.9 s | 97 | Eligible |
Where today's spend goes
32.8%
of today's spend saved
- You pay$501
- Model routing$116
- Context & cache$83
- Language ramps$19
- Clara discount$26
Decision trace: Ticket triage
- Route the request in its original language
- Answer with DeepSeek V4 Pro
Language ramp skipped: on this route it would cost more than it saves
Monthly cost breakdown
- Everything on GPT-6.1 Sol
- $746
- Model routing
- -$116
- Context & cache
- -$83
- Language ramps
- -$19
- Clara discount
- -$26
- With Clara Router
- $501
List prices checked October 27, 2026: standard tier per 1M tokens, without temporary promotions or off-peak discounts. Optimization of up to 40% depends on your workload. The Clara discount applies after optimization, so 40% and 5% combine to 43%.
Ready to route real traffic?
Send us one workload and we'll show what Clara Router would have done with it.