Groq Rate Limits by Tier
What are Groq's API rate limits?
Groq's entry tier (Free) allows 30 requests/min and 6,000 tokens/min for documented default/model family. Limits scale up through Developer as cumulative spend and account age increase — see the full table below, verified 2026-08-14.
Limits by tier
| Tier | Qualification | Model class | RPM | TPM | RPD | Concurrent |
|---|---|---|---|---|---|---|
| Free | Free plan | documented default/model family | 30 | 6,000 | 14400 | — |
| Developer | Developer plan | documented default/model family | 1000 | 250,000 | — | — |
— means not documented by Groq, never a guess.
What this means for your workload
Classification at volume: 116 calls/min and 60,320 tokens/min at the production profile.
At the production volume in classification-at-volume (116 calls/min; 60,320 tokens/min), the documented cap is not sufficient for RPM 30 and TPM 6,000.
Response headers
retry-after | Seconds to wait before retrying, when supplied with a 429 |
rate-limit response headers | Provider-specific remaining and reset counters when documented |
When you exceed the limit
Groq returns HTTP 429.
Respect the model and project limits; cached tokens do not count toward limits.
FAQ
What happens when I exceed Groq's rate limit?
Groq returns HTTP 429 with a retry-after header telling you how long to wait. Respect the model and project limits; cached tokens do not count toward limits.
How do I request a rate limit increase on Groq?
Request one from the account dashboard: https://console.groq.com/settings/limits
Batch 48 · groq rate-limit decision and evidence modules. Surface verification: 2026-08-14. These are dated, route-local fixtures, not live quota claims.
Groq model-plan bucket registry
Frozen groq evidence board: current documented models · Free · Developer · enterprise exception · compound · speech · split input/output limits. Formula / decision rule: bucket fields remain model/plan scoped; unavailable exception cells are not inferred
| Frozen fixture | Inputs and observation | Formula / boundary | State |
|---|---|---|---|
batch48-groq-m1-r1current documented models · Free · Developer | principal/account=joined; endpoint=joined; model/pool=joined; bucket/window=joined; observed=2026-08-14 The primary identity and product joins are present for this frozen fixture. | bucket fields remain model/plan scoped; unavailable exception cells are not inferred A missing provider, product, account, endpoint, model, bucket, or time-window join fails closed; no value is copied across routes. | PASS WITH SCOPE — frozen observation only. |
batch48-groq-m1-r2enterprise exception · compound · speech | request/response/error=joined; limit/remaining/reset=observed; retry identity=joined; region=joined The response evidence is usable only for the named bucket and time window; no neighboring provider is borrowed. | bucket fields remain model/plan scoped; unavailable exception cells are not inferred A missing provider, product, account, endpoint, model, bucket, or time-window join fails closed; no value is copied across routes. | PASS WITH LIMIT — dated evidence only. |
batch48-groq-m1-r3split input/output limits | workload=joined; prompt/output reserve=explicit; shared pool=declared; unknown=Unavailable; owner=platform Admission or recovery is withheld where the account-specific value or side-effect identity is absent. | bucket fields remain model/plan scoped; unavailable exception cells are not inferred A missing provider, product, account, endpoint, model, bucket, or time-window join fails closed; no value is copied across routes. | UNAVAILABLE — live owner evidence required. |
Provenance: Frozen Batch 48 groq fixture: module 1 observation and calculation. First-party evidence checked 2026-08-14; unresolved joins fail closed. Surface verification date: 2026-08-14. Groq rate-limit docs (first-party source)
Groq header-to-headroom decoder
Frozen groq evidence board: normal response · request-daily near-limit · token-minute near-limit · split ITPM/OTPM · audio exhaustion · Retry-After · missing headers. Formula / decision rule: headroom = limit − remaining; daily request units never convert to minute token units
| Frozen fixture | Inputs and observation | Formula / boundary | State |
|---|---|---|---|
batch48-groq-m2-r1normal response · request-daily near-limit | principal/account=joined; endpoint=joined; model/pool=joined; bucket/window=joined; observed=2026-08-14 The primary identity and product joins are present for this frozen fixture. | headroom = limit − remaining; daily request units never convert to minute token units A missing provider, product, account, endpoint, model, bucket, or time-window join fails closed; no value is copied across routes. | PASS WITH SCOPE — frozen observation only. |
batch48-groq-m2-r2token-minute near-limit · split ITPM/OTPM | request/response/error=joined; limit/remaining/reset=observed; retry identity=joined; region=joined The response evidence is usable only for the named bucket and time window; no neighboring provider is borrowed. | headroom = limit − remaining; daily request units never convert to minute token units A missing provider, product, account, endpoint, model, bucket, or time-window join fails closed; no value is copied across routes. | PASS WITH LIMIT — dated evidence only. |
batch48-groq-m2-r3audio exhaustion · Retry-After · missing headers | workload=joined; prompt/output reserve=explicit; shared pool=declared; unknown=Unavailable; owner=platform Admission or recovery is withheld where the account-specific value or side-effect identity is absent. | headroom = limit − remaining; daily request units never convert to minute token units A missing provider, product, account, endpoint, model, bucket, or time-window join fails closed; no value is copied across routes. | UNAVAILABLE — live owner evidence required. |
Provenance: Frozen Batch 48 groq fixture: module 2 observation and calculation. First-party evidence checked 2026-08-14; unresolved joins fail closed. Surface verification date: 2026-08-14. Groq rate-limit docs (first-party source)
Mixed-text-audio admission planner
Frozen groq evidence board: short chat · 32K prompt · 8K output · cached repeat · compound request · transcription hour · 64-request burst. Formula / decision rule: admit = all applicable request/token/audio buckets known and headroom-positive
| Frozen fixture | Inputs and observation | Formula / boundary | State |
|---|---|---|---|
batch48-groq-m3-r1short chat · 32K prompt · 8K output | principal/account=joined; endpoint=joined; model/pool=joined; bucket/window=joined; observed=2026-08-14 The primary identity and product joins are present for this frozen fixture. | admit = all applicable request/token/audio buckets known and headroom-positive A missing provider, product, account, endpoint, model, bucket, or time-window join fails closed; no value is copied across routes. | PASS WITH SCOPE — frozen observation only. |
batch48-groq-m3-r2cached repeat · compound request | request/response/error=joined; limit/remaining/reset=observed; retry identity=joined; region=joined The response evidence is usable only for the named bucket and time window; no neighboring provider is borrowed. | admit = all applicable request/token/audio buckets known and headroom-positive A missing provider, product, account, endpoint, model, bucket, or time-window join fails closed; no value is copied across routes. | PASS WITH LIMIT — dated evidence only. |
batch48-groq-m3-r3transcription hour · 64-request burst | workload=joined; prompt/output reserve=explicit; shared pool=declared; unknown=Unavailable; owner=platform Admission or recovery is withheld where the account-specific value or side-effect identity is absent. | admit = all applicable request/token/audio buckets known and headroom-positive A missing provider, product, account, endpoint, model, bucket, or time-window join fails closed; no value is copied across routes. | UNAVAILABLE — live owner evidence required. |
Provenance: Frozen Batch 48 groq fixture: module 3 observation and calculation. First-party evidence checked 2026-08-14; unresolved joins fail closed. Surface verification date: 2026-08-14. Groq rate-limit docs (first-party source)