carve-webgpu-snowboarding
SHA-256 d720fabc41be9e910c6a74af9d2bcec68e822dea8638eb9e97741c5186cdb1de
Exact prompt file path was not published.
Versioned benchmark collection
Snowboarding
Public ledger · prompt hash recorded · 8 measured runs across 1 target. Each entry is a single recorded attempt; repeat-run reliability is not established.
8Measured runs
1Frozen targets
8Model families
8Configurations
Collection scope
Recorded evidence, without a synthetic score.
Runs share the same frozen target and prompt within each test, with disclosed provider-specific stacks. Costs are recorded estimates as labelled in each ledger, not subscription invoices.
Targets
- CARVE
Model coverage
- Claude Fable 5.1 · Max
- GPT-5.6 Sol · Max
- GPT-6 Astra · Max
- GPT-6 Sol · Max
- Grok 4.6 · xhigh
- MiMo V2.6 Pro · High
- Claude Opus 5 · Max
- Claude Opus 5.5 · Max
Evidence available
- 8 runs on prompt d720fabc41be
- Prompt SHA-256 d720fabc41be9e910c6a74af9d2bcec68e822dea8638eb9e97741c5186cdb1de
- Exact prompt file path was not published
Still missing
- No released Builder project, harness package, or skill download.
- Exact prompt file path was not published with this test.
- 3 related records are listed separately and are not attached to this prompt.
- Repeat-run reliability evidence is not established.
Public receipts
8 measured run ledgers
Open any row for its recorded workflow cost, timing, token usage, stack disclosure, artifact integrity, and explicit evidence gaps.
01Claude Fable 5.1Max reasoning · CARVEpartial_user_turn_count_unverified_token_timing_cost_build_and_operator_webgpu_replay_ledger · Prompt version v1$492.04 Token-only API-equivalent estimate, not the actual Claude Code subscription chargerecorded estimate2h 20m 30.3s wall-clockEnd-to-end workflowOpen 02GPT-5.6 SolMax reasoning · CARVEpartial_two_user_turn_token_timing_source_build_ledger · Prompt version v1$27.40 Token-only standard API-equivalent estimate, not the actual Codex subscription chargerecorded estimate2h 19m 55s wall-clockEnd-to-end workflowOpen 03GPT-6 AstraMax reasoning · CARVEpartial_model_supplied_runtime_evidence_with_independent_clean_build · Prompt version v1$39.12 Standard API-equivalent estimate from the supplied Codex receipt, not an actual subscription charge.recorded estimate1h 52m 02.7s wall-clockEnd-to-end workflowOpen 04GPT-6 SolMax reasoning · CARVEcomplete_token_timing_artifact_and_validation_ledger · Prompt version unreported$13.16 API-equivalent list price, not a marginal subscription cash chargerecorded estimateNot recordedEnd-to-end workflowOpen 05Grok 4.6xhigh reasoning · CARVEpartial_metrics_source_and_production_build_ledger · Prompt version v1$12.66 Provider-recorded usage estimate, not an itemized subscription cash chargerecorded estimate25m 24.387s wall-clockEnd-to-end workflowOpen 06MiMo V2.6 ProHigh reasoning · CARVEPARTIAL_PLAYABLE_GAME_ASSET_QUALITY_UNVERIFIED · Prompt version unreported$0.62 OpenCode-recorded cost, not an itemized subscription cash chargerecorded estimate5539.0s wall-clockEnd-to-end workflowOpen 07Claude Opus 5Max reasoning · CARVEpartial_root_session_metrics_source_and_build_ledger · Prompt version v1$57.60 Root-session token-only standard API-equivalent estimate, not the actual Claude Code subscription charge or a full all-agent campaign costrecorded estimate5h 39m 17.2s wall-clockEnd-to-end workflowOpen 08Claude Opus 5.5Max reasoning · CARVEpartial_gameplay_no_finish_and_asset_judge_fail · Prompt version unreported$57.68 Independently recomputed API-equivalent costrecorded estimateNot recordedEnd-to-end workflowOpen
Prompt provenance
Exact prompt receipts
A path is shown only when the published record includes the prompt hash that identifies it. Campaign evidence and projections keep the hash gap visible.
Separate records
Related records with unknown prompt provenance
These published records stay separate from the prompt receipts above. Unknown provenance is why they are listed here.