#How should I go about removing the 10k character limit when using the API for generating text? I ne

1 messages · Page 1 of 1 (latest)

reef kiteBOT
#
AI-generated Answer for oscarhoh

Short answer: you can’t “remove” the 10k character cap on models that have it. Use a model with a higher per‑request limit, or split your text.

What works today

  • Switch to a v2.5 TTS model. Flash v2.5 or Turbo v2.5 accept up to 40,000 characters per API request (enough for 15–20k). Set model_id to eleven_flash_v2_5 or eleven_turbo_v2_5. (elevenlabs.io)
  • If you must stay on Multilingual v2 (10,000‑char limit), split your input into multiple requests. To keep prosody consistent across chunks, use Request Stitching. (elevenlabs.io)
  • For very long works (chapters/books) with minimal manual splitting, use Studio (formerly Projects). (help.elevenlabs.io)

Reference limits and errors

  • Model‑specific character limits (including 40k for Flash/Turbo v2.5) are documented here and are enforced; exceeding them yields max_character_limit_exceeded. (elevenlabs.io)

If you meant something other than Text‑to‑Speech, let me know which ElevenLabs API you’re using.

Sources:
Models | ElevenLabs Documentation
Models | ElevenLabs Documentation
What's the maximum amount of characters and text I can generate? – ElevenLabs
Models | ElevenLabs Documentation

wooden mountain
#

Additionally, I have some projects where it costs me 7k tokens to generate a 20 minute text, and other projects where the exact same text with exact same voice costs 14k-15k. Where do I check the project settings to make sure I am using a model that costs the least amount of tokens?