SambaNova Cloud 429 RateLimitError on the Developer account

[Coby, SambaNova community team]: The rate limit is the number of API calls within a given minute. If a chat app calls the model API 2-3 times per user request, it gets far fewer user interactions per minute; serving many users on one key needs a queue so requests do not blow past the limit. On the client side, the sambanova SDK retries 429s twice by default with a short exponential backoff, configurable with the maxretries option (set maxretries=0 to disable, or raise it with client.with_options).

Context: Calls to api.sambanova.ai start failing with 429 RateLimitError at what looks like modest volume. SambaNova staff explain the rate limit counts API calls within a given minute per API key, so one app request that fans out to 2-3 model calls burns the budget fast. The free tier is 20 requests per day.

Matched source

Source: Source: https://community.sambanova.ai/t/rate-limit-informations/259 Original query: "SambaNova Cloud 429 RateLimitError on the Developer account" Key terms: account, cloud, developer, ratelimiterror, sambanova

Source: Calls to api.sambanova.ai start failing with 429 RateLimitError at what looks like modest volume. SambaNova staff explain the rate limit counts API calls within a given minute per API key, so one app request that fans out to 2-3 model calls burns the budget fast. The free tier is 20 requests per day.

Verified resolution: [Coby, SambaNova community team]: The rate limit is the number of API calls within a given minute. If a chat app calls the model API 2-3 times per user request, it gets far fewer user interactions per minute; serving many users on one key needs a queue so requests do not blow past the limit. On the client side, the sambanova SDK retries 429s twice by default with a short exponential backoff, configurable with the maxretries option (set maxretries=0 to disable, or raise it with client.with_options).

Source URL: https://community.sambanova.ai/t/rate-limit-informations/259