Groq vs Jina AI

Groq runs open models such as Llama, Mixtral, and Gemma on its own LPU inference hardware, exposing them through a REST API that's a near drop-in replacement for the OpenAI chat completions endpoint. Pricing is pay-per-token and varies by model (roughly $0.05-$3 per million tokens), with a free API key available to get started and no separate infrastructure fee. The console also offers batch processing at a 50% discount and prompt-caching discounts for repeated context.

Jina AI provides token-billed APIs for multilingual, multimodal embeddings, a reranker for search relevance, and the Reader endpoint (r.jina.ai) that converts any URL into clean Markdown for LLM grounding. New accounts get an auto-generated API key with a ten-million-token free trial before moving to paid rate tiers (100 RPM/100K TPM free, up to 5,000 RPM/50M TPM on the top plan). Recent additions include an Elasticsearch-native inference plugin and MCP/CLI integrations for agent workflows.