- Providers
- 1
- Calls, 30 days
- 310K
- Lowest per hit
- $0.0008
- Reviews
- 44
Overview
Hit: A completion with a stop reason.
No provider routes this capability yet.
Endpoints
Every provider of this capability, ranked by Grid score: reliability, fill, accuracy, speed and cost per verified hit, weighted.
| # | Provider | Grid score | Price | Reliability | Fill rate | Accuracy | p50 | Freshness | Per verified hit | Reviews | Last verified |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Rank 1 |  OpenRouter Verified Benchmarked BYOK | 98 | $0.0008per call | 99.8%n=310K | 99.0% | n/a | 1.4 s | live | $0.0008 | Rated 4.7 out of 5(44 reviews) |
Pricing
List price per call, and what one verified hit costs once misses are counted. Failed calls are free.
| Provider | List price | Unit | Per verified hit | Your own key |
|---|---|---|---|---|
| OpenRouter | $0.0008 | per call | $0.0008 | Free, not metered |
Performance
Fill rate, accuracy, success and gateway latency per provider over the trailing 30 days, with the sample behind each number.
| Provider | Fill rate | Accuracy | Success | p50 | p95 | Samples |
|---|---|---|---|---|---|---|
| OpenRouter | 99.0% | n/a | 99.8% | 1.4 s | 6.8 s | 310K |
Uptime
Share of calls without a vendor-side failure (errors, timeouts, capacity). Caller mistakes and refusals don't count against a vendor. Last 30 days, daily.
Sample dataIllustrative figures from the catalog's quality samples until live traffic is published
- OpenRouteropenrouter/ai.complete
- Uptime
- 99.0%
- p50 / p95
- 1.4 s6.8 s
- Calls
- 241.6K
Apps
Public apps that identify themselves with X-Grid-App or HTTP-Referer and X-Title, over the last 30 days.
No public app has listed itself here yet. Add two headers to get yours ranked.
Activity
Routed calls per day over the last 30 days.
- OpenRouter
| Day | OpenRouter |
|---|---|
| Aug 28, 2026 | 8760 |
| Aug 29, 2026 | 4749 |
| Aug 30, 2026 | 4498 |
| Aug 31, 2026 | 10536 |
| Sep 1, 2026 | 8945 |
| Sep 2, 2026 | 9601 |
| Sep 3, 2026 | 9249 |
| Sep 4, 2026 | 9594 |
| Sep 5, 2026 | 4949 |
| Sep 6, 2026 | 5085 |
| Sep 7, 2026 | 10765 |
| Sep 8, 2026 | 10034 |
| Sep 9, 2026 | 10423 |
| Sep 10, 2026 | 10170 |
| Sep 11, 2026 | 8576 |
| Sep 12, 2026 | 4099 |
| Sep 13, 2026 | 4550 |
| Sep 14, 2026 | 9276 |
| Sep 15, 2026 | 10991 |
| Sep 16, 2026 | 9189 |
| Sep 17, 2026 | 8883 |
| Sep 18, 2026 | 9916 |
| Sep 19, 2026 | 4498 |
| Sep 20, 2026 | 4129 |
| Sep 21, 2026 | 8668 |
| Sep 22, 2026 | 10031 |
| Sep 23, 2026 | 8261 |
| Sep 24, 2026 | 9401 |
| Sep 25, 2026 | 9303 |
| Sep 26, 2026 | 4496 |
API
Same input for every provider. Add route=waterfall to fall through on misses, or provider=<slug> to pin one.
| Input | Description |
|---|---|
| model* | Model id (required) |
| prompt* | Prompt or messages (required) |
curl https://api.gridrouter.io/v1/call/ai.complete \
-H "Authorization: Bearer $GRID_KEY" \
-d '{"model":"~anthropic/claude-opus-latest","prompt":"Write a first line.","route":"best_value"}'Fields
The normalized output every provider maps into. Suppliers come from integrated endpoint mappings (verified) or field names in each vendor's API docs (vendor-claimed).
Reviews
No written reviews yet. Reviews open after 100 routed calls on an endpoint.
Questions about this capability
What does a verified model completion result cost?
OpenRouter is the lowest at $0.0008 per verified result once misses and wrong answers are counted. Its list price is $0.0008 per call.
What counts as a hit for model completion?
A completion with a stop reason. Fill rate is the share of calls that produced a hit. Per-hit endpoints bill only hits, and failed calls are never billed.
Can I use my own provider key?
Yes for OpenRouter. Calls on your own key are routed and scored like any other, and never metered.