run-llama/llama_index
120 source files scanned · 37 LLM call site(s) · ~0% of LLM spend looks recoverable (static heuristics · repo looks like a library)
Library/SDK layout detected (Python package "llama-index" is an LLM SDK/framework): call sites are most likely pass-through wrappers where the caller chooses the model and prompt, so findings are downgraded one level and the estimate is discounted.
[](https://tokenzero.sh/audit/run-llama/llama_index)
Models referenced
none detected
Findings
| Severity | Where | Finding | Est. |
|---|---|---|---|
| low | llama-index-integrations/llms/llama-index-llms-google-genai/llama_index/llms/google_genai/base.py:655 | LLM call inside a loop Per-item calls repeat the full prompt prefix for every element. If items are independent and latency-tolerant, the provider Batch API is 50% cheaper, and OpenAI Flex processing (`service_tier: "flex"`) gives the same 50% on ordinary calls without restructuring; if the prefix is shared, caching pays off immediately. Fix: Batch where possible, or set `service_tier: "flex"` on OpenAI calls that can wait; otherwise keep the shared prefix cache-formatted (TokenZero does this automatically). | 25% |
| low | llama-index-integrations/llms/llama-index-llms-google-genai/llama_index/llms/google_genai/base.py:708 | LLM call inside a loop Per-item calls repeat the full prompt prefix for every element. If items are independent and latency-tolerant, the provider Batch API is 50% cheaper, and OpenAI Flex processing (`service_tier: "flex"`) gives the same 50% on ordinary calls without restructuring; if the prefix is shared, caching pays off immediately. Fix: Batch where possible, or set `service_tier: "flex"` on OpenAI calls that can wait; otherwise keep the shared prefix cache-formatted (TokenZero does this automatically). | 25% |
| low | llama-index-integrations/llms/llama-index-llms-google-genai/llama_index/llms/google_genai/base.py:759 | LLM call inside a loop Per-item calls repeat the full prompt prefix for every element. If items are independent and latency-tolerant, the provider Batch API is 50% cheaper, and OpenAI Flex processing (`service_tier: "flex"`) gives the same 50% on ordinary calls without restructuring; if the prefix is shared, caching pays off immediately. Fix: Batch where possible, or set `service_tier: "flex"` on OpenAI calls that can wait; otherwise keep the shared prefix cache-formatted (TokenZero does this automatically). | 25% |
Fix all of it with one line
OPENAI_BASE_URL=https://tokenzero.sh/v1 ANTHROPIC_BASE_URL=https://tokenzero.sh
Keys pass through untouched. Cache formatting and exact caching are on by default; cascade routing stays opt-in. How it works