specification: API Commons Rate Limits specificationVersion: '0.1' schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits provider: ElevenLabs providerId: elevenlabs generated: '2026-09-17' method: searched created: '2026-05-04' modified: '2026-09-17' source: | https://elevenlabs.io/docs/overview/models#concurrency-and-priority (the published per-plan concurrency table and the response headers) and https://elevenlabs.io/docs/eleven-api/resources/errors (429 codes and backoff guidance), both read 2026-09-17. Supersedes the 2026-05-04 generated sweep. docs: https://elevenlabs.io/docs/overview/models description: | ElevenLabs throttles on CONCURRENCY, not on a requests-per-minute quota. The docs say so explicitly: "API requests per minute and concurrent requests are different metrics" and no per-minute ceiling is published. The limit that binds is how many requests a plan may have in flight at once, and it differs by model family (Multilingual v2 vs Flash), by Speech to Text (elevated), by realtime STT and by Music. Beyond the ceiling requests are queued rather than rejected — the docs put the added latency at roughly 50ms — and a 429 is returned when the queue itself is exceeded. model: concurrency tags: - Rate Limiting - Concurrency - Speech headers: request: - name: xi-api-key note: per-key limits; a key can also carry its own credit quota. response: - name: current-concurrent-requests description: how many of your requests are in flight right now. - name: maximum-concurrent-requests description: your plan's concurrency ceiling. retryAfter: null retryAfter_note: Retry-After is NOT documented and was not observed. The provider's guidance is client-side exponential backoff, not a server-supplied delay. standard_ratelimit_headers: none — no X-RateLimit-* and no RFC 9331-style RateLimit-*. responseCodes: throttled: 429 concurrencyExceeded: 429 error_codes: - code: rate_limit_exceeded type: rate_limit_error http_status: 429 remediation: implement exponential backoff before retrying. - code: concurrent_limit_exceeded type: rate_limit_error http_status: 429 remediation: wait for in-flight requests to complete before issuing new ones. priority_queue: behaviour: once the concurrency limit is met, further requests are queued alongside lower-priority requests rather than failing. typical_added_latency_ms: 50 priority_levels: Free: 3 Starter: 4 Creator: 5 Pro: 5 Scale: 5 Business: 5 Enterprise: 6 limit_count: 35 limits: - name: Free — Multilingual v2 concurrency scope: per-account plan: Free metric: concurrent limit: 2 - name: Free — Flash concurrency scope: per-account plan: Free metric: concurrent limit: 4 - name: Free — Speech to Text concurrency scope: per-account plan: Free metric: concurrent limit: 8 - name: Free — Realtime STT concurrency scope: per-account plan: Free metric: concurrent limit: 6 - name: Free — Music concurrency scope: per-account plan: Free metric: concurrent limit: 0 - name: Starter — Multilingual v2 concurrency scope: per-account plan: Starter metric: concurrent limit: 3 - name: Starter — Flash concurrency scope: per-account plan: Starter metric: concurrent limit: 6 - name: Starter — Speech to Text concurrency scope: per-account plan: Starter metric: concurrent limit: 12 - name: Starter — Realtime STT concurrency scope: per-account plan: Starter metric: concurrent limit: 9 - name: Starter — Music concurrency scope: per-account plan: Starter metric: concurrent limit: 2 - name: Creator — Multilingual v2 concurrency scope: per-account plan: Creator metric: concurrent limit: 5 - name: Creator — Flash concurrency scope: per-account plan: Creator metric: concurrent limit: 10 - name: Creator — Speech to Text concurrency scope: per-account plan: Creator metric: concurrent limit: 20 - name: Creator — Realtime STT concurrency scope: per-account plan: Creator metric: concurrent limit: 15 - name: Creator — Music concurrency scope: per-account plan: Creator metric: concurrent limit: 2 - name: Pro — Multilingual v2 concurrency scope: per-account plan: Pro metric: concurrent limit: 10 - name: Pro — Flash concurrency scope: per-account plan: Pro metric: concurrent limit: 20 - name: Pro — Speech to Text concurrency scope: per-account plan: Pro metric: concurrent limit: 40 - name: Pro — Realtime STT concurrency scope: per-account plan: Pro metric: concurrent limit: 30 - name: Pro — Music concurrency scope: per-account plan: Pro metric: concurrent limit: 2 - name: Scale — Multilingual v2 concurrency scope: per-account plan: Scale metric: concurrent limit: 15 - name: Scale — Flash concurrency scope: per-account plan: Scale metric: concurrent limit: 30 - name: Scale — Speech to Text concurrency scope: per-account plan: Scale metric: concurrent limit: 60 - name: Scale — Realtime STT concurrency scope: per-account plan: Scale metric: concurrent limit: 45 - name: Scale — Music concurrency scope: per-account plan: Scale metric: concurrent limit: 5 - name: Business — Multilingual v2 concurrency scope: per-account plan: Business metric: concurrent limit: 15 - name: Business — Flash concurrency scope: per-account plan: Business metric: concurrent limit: 30 - name: Business — Speech to Text concurrency scope: per-account plan: Business metric: concurrent limit: 60 - name: Business — Realtime STT concurrency scope: per-account plan: Business metric: concurrent limit: 45 - name: Business — Music concurrency scope: per-account plan: Business metric: concurrent limit: 5 - name: Enterprise — Multilingual v2 concurrency scope: per-account plan: Enterprise metric: concurrent limit: null limit_note: published as "Elevated"; no number. - name: Enterprise — Flash concurrency scope: per-account plan: Enterprise metric: concurrent limit: null limit_note: published as "Elevated"; no number. - name: Enterprise — Speech to Text concurrency scope: per-account plan: Enterprise metric: concurrent limit: null limit_note: published as "Elevated"; no number. - name: Enterprise — Realtime STT concurrency scope: per-account plan: Enterprise metric: concurrent limit: null limit_note: published as "Elevated"; no number. - name: Enterprise — Music concurrency scope: per-account plan: Enterprise metric: concurrent limit: null limit_note: published as "Highest" priority; no number. notes: - Startup grant recipients receive Scale-level concurrency benefits. - ElevenAgents adds a separate concurrency dimension — simultaneous calls per agent — with burst pricing and call queueing as the documented overflow behaviour (https://elevenlabs.io/docs/eleven-agents/guides/burst-pricing and .../call-queueing). - An individual API key can carry its own credit quota and IP allowlist independent of the plan ceiling (see authentication/elevenlabs-authentication.yml).