Block/goose
120 source files scanned · 2 LLM call site(s) · ~7% of LLM spend looks recoverable (static heuristics)
$80per 10k runs today (gpt-4o, ~2000 prompt tokens)
$74.4per 10k runs with fixes applied
[](https://tokenzero.sh/audit/Block/goose)
Models referenced
gpt-4o ×2 (cheaper sibling: gpt-4o-mini)
Findings
| Severity | Where | Finding | Est. |
|---|---|---|---|
| low | scripts/bench-postprocess-scripts/llm-judges/llm_judge.py:82 | 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 | scripts/bench-postprocess-scripts/llm-judges/llm_judge.py:124 | 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 | scripts/bench-postprocess-scripts/llm-judges/llm_judge.py:82 | "gpt-4o" retires on 2027-01-14 The provider announced on 2026-10-14 that this model retires on 2027-01-14 (80 days). Recommended replacement: gpt-6-luna. Fix: Move to gpt-6-luna (eval first); TokenZero's cascade can route to it, and the gateway warns in X-TokenZero-Warning until you do. | 0% |
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