[ { "slug": "gpt-6-sol", "forecast_run_id": "35776235829", "score_run_id": "35885171489", "model_key": "openai:gpt-6-sol", "covered_case_count": 91, "missing_case_count": 0, "workload_cost_usd": 5.504933, "workload_cost_basis": "estimated_from_pricing_snapshot", "workload_cost_method": "uncached_usage_estimate_missing_cache_usage", "standard_rate_estimate_usd": 11.0097835, "standard_rate_method": "managed_usage_reprice_with_conservative_uncached_fallback", "pricing_reference": "https://developers.openai.com/api/docs/models/gpt-6-sol: Flex is 50% of Standard; rates include cache read (0.1x) and write (1.25x). Above 272000 input tokens, input/cache rates double and output rates multiply by 1.5. Checked 2026-09-22.", "standard_rate_multiplier": 2, "response_count": 771, "missing_cache_dimensions": { "cache_read_tokens": 135, "cache_write_tokens": 44 }, "cache_read_rate_available": true, "cache_write_rate_available": true, "caveats": [ "Successful case receipts only; excludes historical failed attempts, retries, and unresolved charges.", "Missing cache dimensions remain missing; the shared pricing helper uses ordinary-input estimates for affected responses, which may overstate or understate cache-adjusted costs.", "Standard rates are twice the frozen Flex rates, as documented in the frozen registry.", "Receipt totals sum per-case rounded microUSD; standard repricing uses unrounded response costs, so tiny rounding differences remain.", "These prices are workload estimates or reported successful-call charges, not provider invoice totals." ], "pricing_url": "https://developers.openai.com/api/docs/models/gpt-6-sol" }, { "slug": "gpt-6-luna", "forecast_run_id": "35933696453", "score_run_id": "35940612024", "model_key": "openai:gpt-6-luna", "covered_case_count": 91, "missing_case_count": 0, "workload_cost_usd": 0.312309, "workload_cost_basis": "estimated_from_pricing_snapshot", "workload_cost_method": "uncached_usage_estimate_missing_cache_usage", "standard_rate_estimate_usd": 0.62452294, "standard_rate_method": "managed_usage_reprice_with_conservative_uncached_fallback", "pricing_reference": "https://developers.openai.com/api/docs/models/gpt-6-luna: Flex is 50% of Standard; rates include cache read (0.1x) and write (1.25x). Above 272000 input tokens, input/cache rates double and output rates multiply by 1.5. Checked 2026-09-22.", "standard_rate_multiplier": 2, "response_count": 579, "missing_cache_dimensions": { "cache_read_tokens": 128, "cache_write_tokens": 39 }, "cache_read_rate_available": true, "cache_write_rate_available": true, "caveats": [ "Successful case receipts only; excludes historical failed attempts, retries, and unresolved charges.", "Missing cache dimensions remain missing; the shared pricing helper uses ordinary-input estimates for affected responses, which may overstate or understate cache-adjusted costs.", "Standard rates are twice the frozen Flex rates, as documented in the frozen registry.", "Receipt totals sum per-case rounded microUSD; standard repricing uses unrounded response costs, so tiny rounding differences remain.", "These prices are workload estimates or reported successful-call charges, not provider invoice totals." ], "pricing_url": "https://developers.openai.com/api/docs/models/gpt-6-luna" }, { "slug": "claude-opus-5-5", "forecast_run_id": "35953489581", "score_run_id": "35954306953", "model_key": "anthropic:claude-opus-5-5", "covered_case_count": 91, "missing_case_count": 0, "workload_cost_usd": 20.586702, "workload_cost_basis": "estimated_from_pricing_snapshot", "workload_cost_method": "anthropic_cache_aware_usage_reconstruction", "standard_rate_estimate_usd": 20.5866604, "standard_rate_method": "anthropic_cache_aware_usage_reconstruction", "pricing_reference": "https://platform.claude.com/docs/en/about-claude/pricing: Opus 5.5 standard USD4/M input, USD20/M output, USD0.20/M cache read, USD5/M five-minute cache write; no long-context premium. Checked 2026-09-22.", "standard_rate_multiplier": 1, "response_count": 471, "missing_cache_dimensions": {}, "cache_read_rate_available": true, "cache_write_rate_available": true, "caveats": [ "Successful case receipts only; excludes historical failed attempts, retries, and unresolved charges.", "Receipt totals sum per-case rounded microUSD; standard repricing uses unrounded response costs, so tiny rounding differences remain.", "These prices are workload estimates or reported successful-call charges, not provider invoice totals." ], "pricing_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, { "slug": "grok-4-7", "forecast_run_id": "35799524684", "score_run_id": "35885175070", "model_key": "vercel_ai_gateway:spacexai/grok-4.7", "covered_case_count": 91, "missing_case_count": 0, "workload_cost_usd": 11.517165, "workload_cost_basis": "provider_reported", "workload_cost_method": "gateway_reported_charge", "standard_rate_estimate_usd": 32.999826, "standard_rate_method": "managed_usage_reprice_with_conservative_uncached_fallback", "pricing_reference": "https://docs.x.ai/developers/grok-4-7 (USD 2/M input and USD 6/M output standard rates), checked 2026-09-21", "standard_rate_multiplier": 1, "response_count": 629, "missing_cache_dimensions": { "cache_write_tokens": 629, "cache_read_tokens": 3 }, "cache_read_rate_available": false, "cache_write_rate_available": false, "caveats": [ "Successful case receipts only; excludes historical failed attempts, retries, and unresolved charges.", "Missing cache dimensions remain missing; the shared pricing helper uses ordinary-input estimates for affected responses, which may overstate or understate cache-adjusted costs.", "Frozen registry has no cache rates; standard estimate uses ordinary input rates despite reported cache reads, so it is not a cache-adjusted provider charge.", "Receipt totals sum per-case rounded microUSD; standard repricing uses unrounded response costs, so tiny rounding differences remain.", "These prices are workload estimates or reported successful-call charges, not provider invoice totals." ], "pricing_url": "https://docs.x.ai/developers/grok-4-7" }, { "slug": "gpt-4-1", "forecast_run_id": "35954543028", "score_run_id": "36073985227", "model_key": "openai:gpt-4.1-2025-04-14", "covered_case_count": 91, "missing_case_count": 0, "workload_cost_usd": 38.349434, "workload_cost_basis": "estimated_from_pricing_snapshot", "workload_cost_method": "uncached_usage_estimate_missing_cache_usage", "standard_rate_estimate_usd": 38.34943, "standard_rate_method": "managed_usage_reprice_with_conservative_uncached_fallback", "pricing_reference": "https://developers.openai.com/api/docs/models/gpt-4.1: Standard USD 2/M input, 0.50/M cached input, 8/M output; ordinary input pricing for cache creation. Checked 2026-09-24. Flex is not supported.", "standard_rate_multiplier": 1, "response_count": 2100, "missing_cache_dimensions": { "cache_read_tokens": 302, "cache_write_tokens": 2100 }, "cache_read_rate_available": true, "cache_write_rate_available": true, "caveats": [ "Successful case receipts only; excludes historical failed attempts, retries, and unresolved charges.", "Missing cache dimensions remain missing; the shared pricing helper uses ordinary-input estimates for affected responses, which may overstate or understate cache-adjusted costs.", "Receipt totals sum per-case rounded microUSD; standard repricing uses unrounded response costs, so tiny rounding differences remain.", "These prices are workload estimates or reported successful-call charges, not provider invoice totals." ], "pricing_url": "https://developers.openai.com/api/docs/models/gpt-4.1" }, { "slug": "gemini-3-1-pro-preview", "forecast_run_id": "36075557088", "score_run_id": "36169561374", "model_key": "google:gemini-3.1-pro-preview", "covered_case_count": 91, "missing_case_count": 0, "workload_cost_usd": 59.044034, "workload_cost_basis": "estimated_from_pricing_snapshot", "workload_cost_method": "uncached_usage_estimate_missing_cache_usage", "standard_rate_estimate_usd": 59.044026, "standard_rate_method": "managed_usage_reprice_with_conservative_uncached_fallback", "pricing_reference": "https://ai.google.dev/gemini-api/docs/pricing: Standard text input USD 2/M and output USD 12/M up to 200000 input tokens; USD 4/M and 18/M above that threshold. Cached input USD 0.20/M or 0.40/M above 200000 input tokens. Checked 2026-09-24.", "standard_rate_multiplier": 1, "response_count": 1369, "missing_cache_dimensions": { "cache_read_tokens": 529, "cache_write_tokens": 1369 }, "cache_read_rate_available": true, "cache_write_rate_available": false, "caveats": [ "Successful case receipts only; excludes historical failed attempts, retries, and unresolved charges.", "Missing cache dimensions remain missing; the shared pricing helper uses ordinary-input estimates for affected responses, which may overstate or understate cache-adjusted costs.", "Receipt totals sum per-case rounded microUSD; standard repricing uses unrounded response costs, so tiny rounding differences remain.", "These prices are workload estimates or reported successful-call charges, not provider invoice totals." ], "pricing_url": "https://ai.google.dev/gemini-api/docs/pricing" }, { "slug": "gpt-6-1-sol", "forecast_run_id": "36673524599", "score_run_id": "36741462923", "model_key": "openai:gpt-6.1-sol", "covered_case_count": 91, "missing_case_count": 0, "workload_cost_usd": 10.677807, "workload_cost_basis": "estimated_from_pricing_snapshot", "workload_cost_method": "uncached_usage_estimate_missing_cache_usage", "standard_rate_estimate_usd": 10.677807, "standard_rate_method": "uncached_usage_estimate_missing_cache_usage", "pricing_reference": "https://developers.openai.com/api/docs/models/gpt-6.1-sol: Standard USD2/M input, USD10/M output, USD0.10/M cached input, USD2.50/M cache write. Above 272000 input tokens, input and cache rates double and output rates multiply by 1.5. Flex is 50% of Standard. Checked 2026-09-29.", "standard_rate_multiplier": 1, "response_count": 639, "missing_cache_dimensions": { "cache_read_tokens": 128, "cache_write_tokens": 37 }, "cache_read_rate_available": true, "cache_write_rate_available": true, "caveats": [ "Successful case receipts only; excludes historical failed attempts, retries, and unresolved charges.", "Missing cache dimensions remain missing; responses without those counts were priced as ordinary input when the call was made.", "The amount is the sum of estimated_cost_microusd recorded on the receipts during the run.", "These prices are recorded successful-call estimates, not provider invoice totals." ], "pricing_url": "https://developers.openai.com/api/docs/models/gpt-6.1-sol" }, { "slug": "claude-sonnet-5-5", "forecast_run_id": "36673527185", "score_run_id": "36741468019", "model_key": "anthropic:claude-sonnet-5-5", "covered_case_count": 91, "missing_case_count": 0, "workload_cost_usd": 6.992451, "workload_cost_basis": "estimated_from_pricing_snapshot", "workload_cost_method": "anthropic_cache_aware_usage_reconstruction", "standard_rate_estimate_usd": 6.992451, "standard_rate_method": "anthropic_cache_aware_usage_reconstruction", "pricing_reference": "https://platform.claude.com/docs/en/about-claude/pricing: Sonnet 5.5 standard USD2/M input, USD10/M output, USD0.20/M cache read, USD2.50/M five-minute cache write; no long-context premium. Checked 2026-09-29.", "standard_rate_multiplier": 1, "response_count": 310, "missing_cache_dimensions": {}, "cache_read_rate_available": true, "cache_write_rate_available": true, "caveats": [ "Successful case receipts only; excludes historical failed attempts, retries, and unresolved charges.", "The amount is the sum of estimated_cost_microusd recorded on the receipts during the run.", "These prices are recorded successful-call estimates, not provider invoice totals." ], "pricing_url": "https://platform.claude.com/docs/en/about-claude/pricing" } ]