Telnyx Inference – Gemma 4 26B Chat Completions (Large Context)

Choose this endpoint when you need a capable open-source instruction-tuned model (Gemma 4 26B) for large-context chat tasks, prefer OpenAI-compatible APIs for easy drop-in integration, want pay-per-call pricing without a monthly subscription, and your use case involves long multi-turn conversations or documents up to ~12K characters per message. Prefer this over smaller Gemma models (e.g. Gemma 2B) when response quality matters more than cost, and over reasoning-tier models when you do not need chain-of-thought outputs.

• Payment failure or insufficient USDC balance returns 402 Payment Required
• Input messages exceeding 12,000 characters per message or more than 32 messages returns a validation error
• max_tokens exceeding 3072 returns a schema validation error
• Malformed message objects missing required 'role' or 'content' fields return 400 Bad Request
• Network timeouts possible for very long completions near the token limit
• Model unavailability or Telnyx infrastructure issues may return 503 Service Unavailable