> OrcaRouter
OrcaRouter is an LLM gateway offering zero token markup and adaptive routing across 200+ models. It is OpenAI-compatible and supports both pay-as-you-go and subscription plans with a handful of free-tier models.
// pros
- Verifiable operator: Continuum AI Corp (Chicago, USA), with LinkedIn, Crunchbase/PitchBook entries, disclosed funding (~$13.4M) and an arXiv paper — not an anonymous relay
- OpenAI-compatible /v1 API, with /v1/models publicly listable without auth; integration is a base_url swap
- Official MIT-licensed self-hosted build OrcaRouter-Lite (600+ GitHub stars, still receiving commits) plus an official MCP server, so the hosted layer can be bypassed
- A genuine $0-priced free model tier (DeepSeek V4 Flash Free and others), usable without a credit card
- Claims zero token markup with pass-through upstream pricing, funded by Team/Enterprise subscriptions rather than a spread
- Supports openai / anthropic / gemini endpoint shapes simultaneously, plus a browser Playground
// cons
- Fundamentally a commercial aggregation gateway: most models (GPT-5.x, Claude Opus, Gemini 3.x, etc.) are pay-per-token; the free tier only covers a few small/mid models
- Concrete free-tier rate limits (RPM/RPD, per-request size) are published nowhere — only "conservative caps" and "429 when over the limit"
- The site explicitly states that topping up your balance raises free-tier limits, making the free tier a funnel into paid usage rather than a stable entitlement
- Large free-token claims ("1 billion tokens") come mainly from the company's own press releases and social reposts, with no independent verification
- Almost all online visibility is self-issued PR syndicated via PRNewswire/Yahoo/Morningstar plus tool-directory listings; no independent user reviews were found on Linux.do, V2EX, NodeSeek, Reddit or HN
- Very short operating history (GitHub org created 2026-04; the site panel process only started 2026-08-25), so there is no multi-year reliability record
- Probing /api/status returns a one-api / new-api family open-source panel structure (quota_per_unit=500000, announcement_* fields), which sits awkwardly with the "in-house production gateway" marketing — billing and accounts run on an open-source panel
- BYOK requires handing your own upstream provider keys to a third party; custody and liability boundaries must be assessed by you
- Observability, prompt caching and prompt versioning imply request content is retained; privacy and compliance terms need independent review
Treat OrcaRouter as a commercial aggregation gateway with a small free tier, not as a free community relay. Reasonable use: prototype with the free orcarouter/free, deepseek-v4-flash-free or qwen3.8-27b-free models to validate an OpenAI-compatible integration, then pay per token for flagship models or go BYOK with your own upstream keys. Practical advice: confirm your account's actual rate limits and balance-deduction rules in the console before scaling up; do not build a long-term dependency on the free tier, since its limits can change at any time; if using BYOK, supply revocable sub-keys with spending caps rather than a primary key; keep sensitive data off the service given prompt retention and caching; and if you need full control, evaluate the MIT-licensed self-hosted OrcaRouter-Lite. This report is informational risk triage only and is not an endorsement.
community_reputation
Nearly all discoverable information comes from official channels and aggregator directories: PRNewswire/Yahoo/Morningstar syndications of company press releases, listings on AlternativeTo/Toolify/Microlaunch/Cursor Directory, provider docs at Promptfoo and Mastra, official HuggingFace and Apify accounts, and an arXiv paper (OrcaRouter: A Production-Oriented LLM Router with Hybrid Offline–Online Learning). Positive signals: a named, funded legal entity (Continuum AI Corp, Chicago) and an actively maintained open-source repo (603 stars, 76 forks, 22 open issues, commits as recent as August). Negative signal: essentially no independent user reputation — no first-hand reports about stability, billing or abandonment were found on Chinese tech forums (Linux.do, V2EX, NodeSeek) or Reddit/HN, so the advertised free-quota scale and long-term availability cannot be corroborated. Hence a mildly above-neutral score with reduced confidence.
imported analysis — independent third party, not Uprouter editorial · verify before relying on it
Free tier present — value not quantified
This provider offers a free tier, but its terms don’t convert cleanly into a USD figure (e.g. request-count limits or capacity-dependent pools). We refuse to print a misleading zero here. See the note below and the official pricing page for the exact limits.
note: Free-forever 'Hacker' plan: only $0-priced -free models and orcarouter/free endpoint are genuinely free, limited by request rate; probe showed $0 auto-grant. '1B free tokens'/voucher promos lack verifiable terms.
| Plan | Type | Monthly | Input /1M | Output /1M | Note |
|---|---|---|---|---|---|
| Free tier (source-reported) | free | — | — | — | The free "Hacker" plan is advertised as free forever, no credit card, 3 API keys, 0% token markup — but ordinary models still bill against your balance at upstream rates. What is genuinely free are the $0-priced -free models and the orcarouter/free routing endpoint, which the site says are "limited by request rate, not by a fixed pot of calls," while also stating that "topping up your balance raises the free-tier limits." Official and social channels have also promoted "1 billion free tokens," student credits, hackathon grants and voucher drops, but no publicly verifiable eligibility or expiry terms were found; our probe shows a new account quota hint of $0 (no automatic credit grant). Exact RPM/RPD numbers are not published on the site or model pages. |
| Free Tier | free | — | — | — | Includes access to -free suffixed models like deepseek-v4-flash-free. |
Pricing normalized from public sources — always verify with the provider.
// model prices
| Model | Input /1M | Output /1M | Context |
|---|---|---|---|
| orcarouter/free | — | — | — |
| deepseek/deepseek-v4-flash-free | — | — | — |
| qwen/qwen3.8-27b-free | — | — | — |
| tencent/hy3-free | — | — | — |
<a href="https://uprouter.online/s/orcarouter-ai" target="_blank" rel="noopener"><img src="https://uprouter.online/badge/orcarouter-ai.svg" alt="OrcaRouter live status on UPROUTER.ONLINE" height="40"></a>Paste this into your README, docs or status page. The badge shows the same best-effort probe data as the directory, refreshes automatically, and links back to the full OrcaRouter entry — no tracking, no scripts.
$ Does OrcaRouter have a free tier?
Yes — OrcaRouter offers a free tier. Always confirm current limits on OrcaRouter's official pricing page — free tiers change without notice.
$ How much does OrcaRouter cost per 1M tokens?
OrcaRouter does not publish simple per-1M-token pricing (it may bill per request, per output, or via a subscription plan). Check their official pricing page for exact figures.
$ Is OrcaRouter safe to use?
Uprouter rates OrcaRouter as medium risk. Rated medium risk. a commercial aggregator routing to many upstreams with only partial transparency: some model coverage or pricing is confirmed, while other figures rest on third-party reports or a login-gated console. Risk ratings are editorial, evidence-based, and never influenced by affiliate relationships.
$ Can I use OrcaRouter through Uprouter Connect?
Yes — OrcaRouter is Connect-compatible. Add your OrcaRouter API key in Uprouter Connect (encrypted at rest) and route requests to it through one OpenAI- and Claude-compatible API, optionally behind failover aliases. Uprouter meters the traffic in compute at the provider's listed rate.
Raw community sentiment, shown unblended. Votes never feed into objective data fields like price, risk, or status — and never into sort order.
No published reviews yet — be the first.
// more_informationprice history · live status · risk & confidence · code snippets
price_history
No recorded changes yet.
live_status
full history →best-effort probes · not a guarantee · last 13h ago
| Checked | Status | Latency |
|---|---|---|
| 13h ago | up | 152ms |
| 14h ago | up | 286ms |
| 1d ago | up | 209ms |
| 2d ago | up | 235ms |
| 3d ago | up | 278ms |
| 4d ago | up | 1267ms |
| 12d ago | up | 1972ms |
Uprouter runs its own probes; best-effort, not a guarantee. Probe results depend on our network location and cadence — for production, always use the provider’s own status page.
risk_&_confidence
Rated medium risk. a commercial aggregator routing to many upstreams with only partial transparency: some model coverage or pricing is confirmed, while other figures rest on third-party reports or a login-gated console.
Confidence reflects how much verifiable evidence (official docs, probe history, community reports) backs this entry. Risk is our editorial assessment of operator reliability and terms — not financial advice.
code_snippets
curl https://uprouter.online/api/connect/v1/chat/completions \ -H "Authorization: Bearer upr_live_YOUR_KEY" \ -H "Content-Type: application/json" \ -d '{
"model": "orcarouter-free",
"messages": [{ "role": "user", "content": "Hello via OrcaRouter" }]
}'Metered in Uprouter compute — the usage log shows which upstream served each request and its exact cost.
Ranked by neutral similarity signals only — live status, Connect compatibility, free-tier shape and free-credit proximity. Commercial relationships never influence this list.