Loading
90 days of the enterprise, pro and team plans free, no card needed. See the plans
Cost estimator
Paste a prompt. The router estimates tokens, scores every target and quotes each one with its price split into input and output, expected quality and latency, then names the recommended, cheapest capable, best-quality and fastest picks.
Nothing is executed and the prompt is not stored. Tokens are estimated from the text; costs use each model's list price.
Llama 3.3 70B Instruct
$0.6750 / month at 10K requests
Llama 3.3 70B Instruct
$0.6750 / month at 10K requests
Llama 3.3 70B Instruct
$0.6750 / month at 10K requests
Llama 3.3 70B Instruct
$0.6750 / month at 10K requests
35 tokens in, 200 out (predicted); confidence 29%.
Ranked by the router's utility; cost splits input and output tokens at list price. The routing quality score is the router's own estimate for this prompt (its prior for the model, adjusted by learned outcomes), on a 0-1 scale - not an independent benchmark.
| # | Target | Input | Output | Per request | Routing quality | Latency |
|---|---|---|---|---|---|---|
| 1 | domain fit 100% · complexity fit 100% · quality prior 0.70 | $0.0000035 | $0.0000640 | $0.0000675 | 0.74 | 1.20 s |
| 2 |  domain fit 50% · complexity fit 100% · quality prior 0.55 | $0.0000035 | $0.0000800 | $0.0000835 | 0.64 | 300.0 ms |
| 3 |  domain fit 50% · complexity fit 100% · quality prior 0.60 | $0.0000017 | $0.0000200 | $0.0000218 | 0.64 | 700.0 ms |
| 4 | domain fit 50% · complexity fit 86% · quality prior 0.78 | $0.0000006 | $0.0000180 | $0.0000186 | 0.62 | 1.80 s |
| 5 | domain fit 50% · complexity fit 100% · quality prior 0.75 | $0.0000140 | $0.00032 | $0.00033 | 0.64 | 900.0 ms |
| 6 | rule 'support-domain' prefers · domain fit 50% · action fit 0% | $0.0000700 | $0.00040 | $0.00047 | 0.62 | 1.50 s |
| 7 | domain fit 50% · complexity fit 64% · quality prior 0.93 | $0.0000700 | $0.00160 | $0.00167 | 0.60 | 2.50 s |
| 8 | domain fit 100% · action fit 0% · complexity fit 100% | $0.0000035 | $0.0000200 | $0.0000235 | 0.52 | 1.50 s |
| 9 | Translation skillskill domain fit 50% · action fit 0% · complexity fit 100% | $0.0000175 | $0.00010 | $0.00012 | 0.47 | 400.0 ms |
| 10 | domain fit 100% · action fit 0% · complexity fit 100% | $0.00017 | $0.00100 | $0.00118 | 0.51 | 12.0 s |
| 11 | Summarization skillskill domain fit 50% · action fit 0% · complexity fit 100% | $0.0000280 | $0.00016 | $0.00019 | 0.46 | 1.20 s |
| 12 | domain fit 0% · complexity fit 100% · quality prior 0.84 | $0.00028 | $0.00160 | $0.00188 | 0.49 | 9.00 s |
| 13 | Code modelllm domain fit 0% · action fit 0% · complexity fit 100% | $0.0000035 | $0.0000200 | $0.0000235 | 0.41 | 1.20 s |
| 14 | Customer notification workflowworkflow domain fit 50% · action fit 0% · complexity fit 100% | $0 | $0 | $0.00400 | 0.47 | 2.00 s |
| 15 | Research agentagent domain fit 50% · action fit 0% · complexity fit 79% | $0.00021 | $0.00120 | $0.00141 | 0.46 | 8.00 s |
| 16 | B Web searchtool domain fit 50% · action fit 0% · complexity fit 100% | $0 | $0 | $0 | 0.40 | 1.20 s |
| 17 | B Calculatortool domain fit 0% · action fit 0% · complexity fit 100% | $0 | $0 | $0 | 0.37 | 5.0 ms |
| 18 | SQL skillskill domain fit 0% · action fit 0% · complexity fit 100% | $0.0000350 | $0.00020 | $0.00023 | 0.39 | 700.0 ms |
| 19 | Coding agentagent domain fit 0% · action fit 0% · complexity fit 94% | $0.00028 | $0.00160 | $0.00188 | 0.40 | 6.00 s |
| 20 | Image generatortool domain fit 0% · action fit 0% · complexity fit 100% | $0.00140 | $0.00800 | $0.00940 | 0.37 | 9.00 s |
| 21 | H domain fit 100% · action fit 0% · complexity fit 100% | $0.0175 | $0.1000 | $0.1175 | 0.42 | 5.0 min |
Recommended models
Live quotes of representative prompts against this deployment's catalogue, recomputed every minute. Rankings follow the targets, list prices and learned quality, not a hand-maintained list.
GPT-4.1 Minillm
Why: domain fit 50% · complexity fit 100% · quality prior 0.75
this pickthis pickthis pickcoding-agentagent
Why: rule 'hard-coding-to-agent' prefers · domain fit 100% · action fit 50%
Frontier reasoning model$15/1MFrontier reasoning model$15/1MGPT-4.1 Minillm
Why: domain fit 50% · complexity fit 100% · quality prior 0.75
gpt-oss-20b (self-hosted)$0.40/1Mthis pickthis pickskill-summarizeskill
Why: rule 'legal-persona' prefers · domain fit 100% · complexity fit 100%
On-prem private model$1.00/1MOn-prem private model$1.00/1Mqwen3-vl-4b-instructllm
Why: domain fit 33% · complexity fit 100% · quality prior 0.74
Small fast model$0.20/1MLlama 3.3 70B Instructllm
Why: rule 'pii-stays-onprem' prefers · domain fit 100% · complexity fit 100%
this pickthis pickthis pickrouter's pick for 2 use cases · cheapest 1x · best quality 2x
router's pick for 1 use case · cheapest 2x · best quality 1x
router's pick for 1 use case · cheapest 1x · best quality 1x
router's pick for 1 use case · best quality 1x
router's pick for 1 use case · best quality 1x
2 placements · cheapest 1x
Photo: Jakub Zerdzicki
The same quote is one call away: anonymous quotes are rate limited per client, keyed quotes apply your tenant rules and show up in your usage.
curl -s /api/v1/estimate -H "Authorization: Bearer $OSR_KEY" \
-H "Content-Type: application/json" \
-d '{"text": "Summarise this contract...", "monthly_requests": 50000}'osr estimate quotes against a local catalogue and the MCP server exposes estimate and recommend tools to VS Code, Cursor and Claude.
from opensmartroute.estimate import estimate
q = estimate(router, "Summarise this contract...", output_tokens=300)
print(q.recommended.name, q.recommended.cost_usd, q.savings_usd)Set up the IDE integration under Integrations or read the MCP guide.