Recalculate Gemini-3.5-Flash swe-bench costs ($3.8253/instance)

[all-hands-bot]

🧮 Cost Recalculation for swe-bench

Archive URL: https://results.eval.all-hands.dev/swebench/litellm_proxy-gemini-3-5-flash/26622025994/results.tar.gz

Pricing Notes

• ✅ cache_read_price from metadata.json was used for cache hit costs.
• ℹ️ cache_write_price was ignored in this calculation.

Token Usage Summary

Metric	Value
Instances processed	500
Total prompt tokens	1,439,426,049
Cache read tokens	315,611,980
Non-cached prompt tokens	1,123,814,069
Completion tokens	19,953,716

Pricing Used (from metadata.json)

Token Type	Cost per Million
Input (cache miss)	$1.5000
Input (cache hit)	$0.1500
Output	$9.0000

Cost Breakdown

Component	Tokens	Rate	Cost
Non-cached input	1,123,814,069	$1.5000/M	$1685.7211
Cached input	315,611,980	$0.1500/M	$47.3418
Output	19,953,716	$9.0000/M	$179.5834
Total			$1912.6463

Final Result

• Total cost: $1912.6463
• Average cost per instance: $3.8253

---

This comment was automatically generated by the recalculate-costs action.

[all-hands-bot]

🧮 Cost Recalculation for swe-bench

Archive URL: https://results.eval.all-hands.dev/swebench/litellm_proxy-gemini-3-5-flash/26622025994/results.tar.gz

Pricing Notes

• ✅ cache_read_price from metadata.json was used for cache hit costs.
• ℹ️ cache_write_price was ignored in this calculation.

Token Usage Summary

Metric	Value
Instances processed	500
Total prompt tokens	1,439,426,049
Cache read tokens	315,611,980
Non-cached prompt tokens	1,123,814,069
Completion tokens	19,953,716

Pricing Used (from metadata.json)

Token Type	Cost per Million
Input (cache miss)	$1.5000
Input (cache hit)	$0.1500
Output	$9.0000

Cost Breakdown

Component	Tokens	Rate	Cost
Non-cached input	1,123,814,069	$1.5000/M	$1685.7211
Cached input	315,611,980	$0.1500/M	$47.3418
Output	19,953,716	$9.0000/M	$179.5834
Total			$1912.6463

Final Result

• Total cost: $1912.6463
• Average cost per instance: $3.8253

---

This comment was automatically generated by the recalculate-costs action.