AI Perks curates and provides access to exclusive discounts, credits, and deals on AI tools, cloud services, and APIs to help startups and developers save money.
What are the OpenRouter rate limits on free models?
Free models are capped at 20 requests per minute and 50 per day. Buying at least $10 of credits raises the daily cap to 1,000. Paid models have no plan-based rate limit.
If you are staring at a 429, the fastest permanent fix is usually credits, not code. Up to $3,000 in OpenRouter credits is listed in the AI Perks catalogue, and the card shows who qualifies and how to claim it.
Two details catch people out. Failed attempts count toward the daily free-model cap, and a negative balance blocks free models until you add credits.
Top AI Credits for Startups
Apply directly through these verified programs.
OpenRouter Up to $3,000 + 0% fees for 12 months
Eligible for early-stage startups
Claude (via program) $100,000
Eligible for early-stage startups
OpenAI (via program) $2,500
Eligible for early-stage startups
Anthropic $20,000
Eligible for early-stage startups
AWS $300,000
Eligible for early-stage startups
Google Cloud $350,000
Eligible for early-stage startups
How do free and paid limits compare?
The plan table is simple: the Free plan gets 50 requests per day, paid plans pass through the provider's limits, and Enterprise can get higher allowances.
| Setup | Rate limit |
|---|---|
| Free models, no credits bought | 20 requests/min, 50 requests/day |
| Free models, at least $10 bought | 20 requests/min, 1,000 requests/day |
| Paid models | No plan-based limit, provider limits pass through |
| Enterprise | Higher allowances available |
The free-limit tier is selected by all-time credits purchased, not by the is_free_tier flag. To absorb rounding and top-up fees, the higher daily ceiling is granted starting one credit below the $10 threshold.
For the model list itself, see OpenRouter Free Models: List and Limits.
What does an OpenRouter 429 actually mean?
A 429 has three possible sources, and the fix depends on which one you hit.
1. OpenRouter itself. You hit a platform limit: free-model requests per minute or per day, or DDoS protection. That protection blocks requests that dramatically exceed reasonable usage.
2. The upstream provider. The provider serving your request is rate limiting or at capacity. When available, error.metadata.provider_code carries the provider's original error code. Fallback routing retries other providers for the same model before this error reaches you.
3. Your own provider account (BYOK). If you use your own provider keys, a 429 means the provider account hit its rate limit. Check your provider's settings or wait.
When every attempted provider returned a retry hint, the response also carries a Retry-After header.
Top AI Credits for Startups
Apply directly through these verified programs.
OpenRouter Up to $3,000 + 0% fees for 12 months
Eligible for early-stage startups
Claude (via program) $100,000
Eligible for early-stage startups
OpenAI (via program) $2,500
Eligible for early-stage startups
Anthropic $20,000
Eligible for early-stage startups
AWS $300,000
Eligible for early-stage startups
Google Cloud $350,000
Eligible for early-stage startups
How do you check your key and daily usage?
Make a GET request to the key endpoint at https://openrouter.ai/api/v1/key. It shows the rate limit and credits left on that key, plus the free-model requests recorded so far in the current UTC day.
The per-minute limit is not reported there. Successful responses also carry no X-RateLimit-* headers, which makes the key endpoint your only way to monitor quota.
How do you fix a 429 on OpenRouter?
Match the fix to the cause:
- Retry with exponential backoff. Rate limits are transient, so wait rather than re-sending at once, and honor
Retry-Afterwhen present. - On free variants, buy at least $10 of credits to raise the daily limit, or switch to the paid variant, which has no platform-level request cap.
- Spread load across models. Extra accounts or API keys do not raise limits, but different models have different limits.
- Do not create extra accounts. The Terms of Service prohibit multiple accounts to get around use limits.
- Use BYOK so limits are managed by your provider account. By default, OpenRouter falls back to shared endpoints if all your keys hit a limit or fail.
Be careful with fallback lists: adding a paid model after a :free model bills the paid model at its own price.
What does the $10 unlock really cost?
The minimum purchase is $5 per transaction. Non-crypto purchases cost 5.5% with a $0.80 minimum fee, and 5.5% of $10 is less than $0.80, so the minimum applies on a $10 top-up. Unused credits can be refunded only within 24 hours of the transaction.
Top AI Credits for Startups
Apply directly through these verified programs.
OpenRouter Up to $3,000 + 0% fees for 12 months
Eligible for early-stage startups
Claude (via program) $100,000
Eligible for early-stage startups
OpenAI (via program) $2,500
Eligible for early-stage startups
Anthropic $20,000
Eligible for early-stage startups
AWS $300,000
Eligible for early-stage startups
Google Cloud $350,000
Eligible for early-stage startups
Which free credits remove the shared free-pool ceiling?
Vendor credits move you off the shared free pool. Paid models carry no plan-based limit, and direct provider credits put you under your own provider's limits. Each card shows who qualifies and how to claim.
| Company | Free credits | How to get |
|---|---|---|
| OpenRouter | Up to $3,000 | Get this perk |
| Anthropic API | $20,000 | Get this perk |
| Claude via partners | Up to $100,000 | Get this perk |
| Google Cloud (Vertex AI, Gemini) | $350,000 | Get this perk |
| AWS (Amazon Bedrock) | $300,000 | Get this perk |
| Azure OpenAI | $150,000 | Get this perk |
| Together AI | Up to $50,000 | Get this perk |
| Cerebras | Up to $22,500 | Get this perk |
| Mistral AI | $10,000 | Get this perk |
| OpenAI API | $1,200 | Get this perk |
Total across these ten perks: $3,000 + $20,000 + $100,000 + $350,000 + $300,000 + $150,000 + $50,000 + $22,500 + $10,000 + $1,200 = up to $1,006,700 in free credits.
BYOK is worth knowing about here. Standard and Business plans include $25,000 per month of list-price inference through BYOK with no OpenRouter fee; above that, OpenRouter charges 5% deducted from your credits. More on routing in Free OpenRouter Credits: Up to $3,000 Plus BYOK.
How do you get out of the free-pool limit step by step?
- Get credits via AI Perks. Open the catalogue at getaiperks.com, pick a card from the table above, and follow the claim steps shown there.
- Check your key. Call the key endpoint and note your credits and the free-model requests used in the current UTC day.
- Switch off
:freefor production traffic. Point the request at the paid variant of the same model, which has no platform-level request cap. - Add backoff. Retry with exponential backoff, honor
Retry-After, and cap retries so a failing loop cannot drain a free quota. - Move heavy traffic to BYOK if you hold credits with a direct provider, so your provider account sets the limits.
For a deeper look at what the free tier allows, read OpenRouter Free Tier: Real Limits, Fusion, and the Stripe Deal.
Top AI Credits for Startups
Apply directly through these verified programs.
OpenRouter Up to $3,000 + 0% fees for 12 months
Eligible for early-stage startups
Claude (via program) $100,000
Eligible for early-stage startups
OpenAI (via program) $2,500
Eligible for early-stage startups
Anthropic $20,000
Eligible for early-stage startups
AWS $300,000
Eligible for early-stage startups
Google Cloud $350,000
Eligible for early-stage startups
FAQ
How many requests can I make to OpenRouter free models per day?
Without any credit purchase you get 50 free-model requests per day, plus a 20 per minute cap. After buying at least $10 of credits, the daily limit becomes 1,000. Failed attempts count too. Check the OpenRouter card for credits that cover paid models.
Do failed requests count against the OpenRouter daily limit?
Yes. Failed attempts count toward the daily free-model cap, so a misconfigured retry loop can use up your allowance before returning anything useful. Errored attempts are not charged for model tokens, but they still burn free-tier quota. Cap your retries and use backoff.
Will a second account or API key raise my limit?
No. Extra accounts or API keys do not raise rate limits, and the Terms of Service prohibit creating multiple accounts to bypass use limits. Different models do have different limits, so spreading requests across models helps. For real headroom, browse the credits at getaiperks.com.
Why do I get a 429 on a paid model?
Paid models have no plan-based OpenRouter limit, so the 429 is usually the upstream provider rate limiting or at capacity. Look at error.metadata.provider_code and any Retry-After header. With BYOK, it means your own provider account hit its limit.
Can I see my per-minute usage?
No. The key endpoint reports credits left and free-model requests used in the current UTC day, but not the per-minute count. Successful responses carry no X-RateLimit-* headers either. Track per-minute calls in your own client, and compare vendor credit options at getaiperks.com.
Ready to stop hitting the free cap?
A 429 on a free model is a pool limit, not a bug in your code. Credits move you onto paid models and your own provider limits, and the cards show exactly who qualifies and how to claim.
Get free AI credits at getaiperks.com →
Browse the full catalogue at getaiperks.com and start your next trial without the shared-pool ceiling.
