Measured harness ledgerPublic result
GPT-6 AstraHigh-voxel-density Oasis Outpost diorama — GPT-6 Astra Max
Create a complete inspectable Blender voxel diorama of an oasis outpost at a fixed 0.05-unit voxel resolution through the Blender MCP workflow.
Max reasoningHeadline result
- Workflow cost
- $10.43
- Wall-clock
- 30m 09.5s (recovered) wall-clock
- Processed tokens
- 7.17M processed
- Record state
- partial_recovered_timing_with_blender_scene_open_inspection
Public summary
GPT-6 Astra Max partial_recovered_timing_with_blender_scene_open_inspection ledger: 30m 09.5s (recovered) wall-clock, 7.17M processed, and $10.43 Standard API-equivalent estimate from the supplied Codex receipt, not an actual subscription charge..
Run identity and stack
- Result ID: oasis-outpost-gpt-6-astra-max
- Technical model: gpt-6-astra
- Provider: OpenAI Codex
- Stack: OpenAI Codex
- Stack: Technical model/configuration: gpt-6-astra
- Stack: Blender MCP
- Stack: 0.05-unit voxel grid
- Stack: Inspectable .blend artifact
Primary artifact integrity
- Kind: Blender scene
- Path: artifacts/oasis-outpost-gpt-6-astra-max/Oasis_Outpost.blend
- SHA-256: 5ad7c465b4f6a5d7f37fbfb746f484f4778b42ad466ecc083be17113b42d5d01
Recorded caveats
- The unchanged parser found no qualifying user timestamp. The separately labelled 1809.5-second wall-clock was recovered from response-item timestamps and is end-to-end latency, not model-only compute.
- Output tokens include hidden reasoning, visible prose/code, and tool-call JSON.
- Cache reads are discounted, so total processed-token volume overstates effective cost.
- The Blender scene was independently opened without rendering or modification; that does not establish visual quality.
- Supplied viewport preflight and geometry audit are retained as supplied evidence. Remote final-render acceptance and blind evaluation remain unverified.
Visible evidence gaps
- independent remote final-render acceptance frames with renderer, resolution, and hardware receipt
- blind-evaluation record
Public result only
This result keeps its public summary and evidence, but it does not currently have a matching Builder test with prompts, projects, Harness workflows or skills.