Telnyx Inference Chat Completions (Gemma 2B)

Choose this endpoint when you need a low-cost ($0.002/call), pay-per-use LLM inference call without a subscription, specifically against the Gemma-2B-IT instruction-tuned model. It is ideal for lightweight tasks like classification, short-form generation, or FAQ responses where a smaller model suffices. Prefer it over larger model endpoints when cost efficiency matters and you are already operating in an x402 micropayment environment. Not suitable when you need longer outputs (>256 tokens), larger context windows, or more capable models.

• Payment failure: x402 payment not fulfilled or insufficient USDC balance causes rejection before inference runs
• Content too long: messages exceeding 2000 characters per message or more than 32 messages returns a validation error
• max_tokens out of range: values below 1 or above 256 are rejected
• Temperature out of range: values outside 0-2 cause schema validation failure
• Model mismatch: specifying any model other than google/gemma-2b-it returns an error since the route is fixed
• Network timeout: inference latency spike causes request to time out with no completion returned