# RightNow System Zero (slim)

> RightNow System Zero is a hosted, closed-weight decision model from RightNow AI, Inc. It answers boolean, choice and score questions about text, JSON, images, WAV audio and PDFs with per-option probabilities, over HTTP or a WebSocket stream.

- Full: https://www.anchorterminal.com/tools/rightnow-system-zero.md (~8,850 tokens) · this version ~1,730 tokens · JSON https://www.anchorterminal.com/tools/rightnow-system-zero.json · canonical https://www.anchorterminal.com/tools/rightnow-system-zero
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-10-31 · pricing https://www.anchorterminal.com/tools/rightnow-system-zero#pricing · reviews https://www.anchorterminal.com/tools/rightnow-system-zero#reviews
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

**B · 66.8/100 · rank #320 of 1244 · #9 in Decision models · not agent-ready · confidence medium**


Assessment: A well-documented decision API with an OpenAPI file, idempotent retries, a detailed error catalogue, streaming over WebSocket and published per-unit prices with USD 5 of starting credit. Standard context is 38,000 tokens; the 64k, 128k and 1M bands are held and video is a preview. The latency and benchmark figures are RightNow's own, and several of its pages disagree.

## Facts

- Kind: Model API · vendor: RightNow AI, Inc. · category: Decision models · legal entity: RightNow AI, Inc. · provenance 84/100
- Endpoint: `https://api.rightnowai.co/v1/system-zero/decide` (HTTP, websocket)
- Auth: API key · pricing: Pay per use · x402: no · licence: Proprietary hosted model under RightNow AI terms of use; the SDKs are Apache-2.0
- Probe metrics: not measured yet (probes haven't run)
- Surface graded: The hosted API at https://api.rightnowai.co/v1 (POST /system-zero/decide, GET /models, GET /credits and the WebSocket stream at /system-zero/stream), from the docs, OpenAPI file and SDK repository read on 31 October 2026. No call was made with a key and no WebSocket was opened.
- Question types: boolean, choice (1 to 255 options) and score (2 to 10 levels), up to 1,024 named questions a call. Each answer has decision, confidence, alternatives, calibrated (false in this release) and calibration_version.
- Inputs served: Text and JSON state. Up to 4 PNG or JPEG images, one 16,000 Hz WAV of up to 300 seconds and PDFs of up to 500 pages, inline only. 2,621,440 decoded bytes and 3,665,920 request bytes a call.
- Preview and held: Video is a preview, one silent H.264 MP4 of up to 10 seconds with no other file. The 64k and 128k bands are held. The 1M context window is held and opt-in from the console Danger zone, at USD 0.87 per million tokens once its pool opens.
- Context: 38,000 state tokens and 40,960 rendered tokens on the served standard band. An over-long state is refused with 413 state_too_large.
- Rate limits: 600 requests a minute per key, 16 calls in flight and 100,000 model calls a UTC day per workspace, 5,000,000 decisions a UTC month. The stream allows one socket per account.
- SDKs: @rightnowai/system-zero on npm and rightnow-system-zero on PyPI, 0.4.2 of 27 October 2026, Apache-2.0. The npm package also exports function tools and an MCP server adapter.
- Data handling: No training on content or answers and no content logging, per the privacy policy and terms. Answers are kept for up to 24 hours only for requests with an Idempotency-Key. Processing can run in any region of the cloud host.
- Vendor claims: 30 to 33 ms median text latency and a 63.20 Decision Index score are RightNow's own measurements. The index score is labelled RightNow 3.0.0.
- Prices: Uncached text input $0.19 per 1M tokens; Cached text input $0.019 per 1M tokens; Answered question $0.0001 per transaction; Answered call fee $0.0001 per call; Image input $0.011 per image; Audio input $0.022 per minute of audio; PDF input $11 per 1,000 pages; Video input (preview) $0.0018 per second of video

- Scores: Reliability 50, Performance pending, Schema & documentation 83, Agent ergonomics 86, Security & auth 70, Payments & pricing 40, Task success pending, Maintenance & community 68, Transparency & trust 70 · total over the 7 assessed categories
- Why: Reliability, Scored as a hosted model API, as for Celeris-1 Decision and Clef. · Schema & documentation, An OpenAPI 3.1.0 file in the official SDK repository covers decide, the model list and the stream, with typed questions and attachments and… · Agent ergonomics, Read as an API an agent calls for a decision, as with the other decision models. · Security & auth, Read for a model API. · Payments & pricing, No x402, MPP or L402. HTTP 402 means the workspace's prepaid balance is too low (0). · Maintenance & community, Read for a model. · Transparency & trust, Closed weights under terms of use from RightNow AI, Inc.
- Sources: 27, open questions: 13, both in the full twin
- Capabilities: inference.decision
- JSON: https://www.anchorterminal.com/api/v1/tools/rightnow-system-zero.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/rightnow-system-zero.svg` or a link to https://www.anchorterminal.com/tools/rightnow-system-zero from a page on rightnowai.co or one of its subdomains, or the README of github.com/RightNow-AI/system-zero-sdk, then `POST https://www.anchorterminal.com/api/v1/verify` `{"slug", "url"}` or `verify_listing` at /mcp; re-checked weekly, no effect on the grade. Snippets in the full twin.
- Support: Discord https://discord.gg/sSJqgNnq6X; https://www.rightnowai.co/docs/system-zero/support; https://www.rightnowai.co/docs/system-zero/errors; more in the full twin

## Before you call it

1. Send audio only as inline 16,000 Hz WAV. The FLAC, MP3, OGG and MP4 audio named on the blog are refused with media_unsupported.
2. Keep state under 38,000 tokens. Long-context requests return 404 long_context_not_available, which never succeeds on retry.
3. Reuse the same Idempotency-Key and body on retries, and stop when error.retryable or x-should-retry is false.
4. Check every entry in answers even on HTTP 200, since one question can be refused while the others are answered.
5. Treat probabilities as uncalibrated while calibrated is false, and pin system-zero-2.0.0 when answers must not move with the alias.

## Connect

```bash
npm install @rightnowai/system-zero@0.4.2
```

```bash
curl https://api.rightnowai.co/v1/system-zero/decide \
  -H "Authorization: Bearer $SYSTEM_ZERO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"system-zero","state":"The blender I ordered arrived with a cracked jar.","questions":{"damaged":{"type":"boolean","instructions":"Did the item arrive damaged?"}}}'
```

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| OpenAI Decisions API | BB | 73.1 | inference.decision | https://www.anchorterminal.com/tools/openai-decisions-api.min.md |
| Decider | BB | 70 | inference.decision | https://www.anchorterminal.com/tools/decider.min.md |
| Drex 1.5 | B | 69.5 | inference.decision | https://www.anchorterminal.com/tools/nace-drex.min.md |
| Laya | B | 69.5 | inference.decision | https://www.anchorterminal.com/tools/convai-laya.min.md |
| pplx-decider | B | 68.1 | inference.decision | https://www.anchorterminal.com/tools/pplx-decider.min.md |


## Panel reviews (0, desk reviews from public material, no calls made)