Measured harness ledgerPublic result
GLM 5.3 FlashHigh-voxel-density Oasis Outpost diorama — GLM 5.3 Flash 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
- $0.23
- Wall-clock
- 35m 0.633s wall-clock
- Processed tokens
- 442K processed
- Record state
- partial_multi_user_turn_artifact_ledger
Public summary
GLM 5.3 Flash Max partial_multi_user_turn_artifact_ledger ledger: 35m 0.633s wall-clock, 442K processed, and $0.23 OpenCode Go API-equivalent usage accounting, not an itemized subscription marginal-cash invoice.
Run identity and stack
- Result ID: oasis-outpost-glm-5.3-max
- Technical model: glm-5.3-flash
- Provider: opencode-go
- Client: OpenCode
- Stack: opencode-go
- Stack: OpenCode
- Stack: Technical model/configuration: glm-5.3-flash
- Stack: Blender MCP
- Stack: 0.05-unit voxel grid
- Stack: Inspectable .blend artifact
Primary artifact integrity
- Kind: blender-scene
- Path: artifacts/oasis-outpost-glm-5.3-max/oasis_outpost.blend
- SHA-256: e760cb31917939756854a924ef61ba2d5111d50ce33027a5d4370bf56a1e3d55
Validation evidence
- Path: artifacts/oasis-outpost-glm-5.3-max/validation/scene-inspection.public.json
- SHA-256: 934d8de69b57f43938dc151d0bec08fa4acc25038a497a2e92a380e7966ac282
Recorded caveats
- This is a six-user-turn recovery record, not a strict one-shot benchmark result.
- The metrics sidecar is operator-generated from the canonical bounded extractor after the same-session metrics stream stalled; the underlying raw records remain private.
- No executable model-authored Blender builder was delivered; the source record is limited to embedded build notes and the editable final scene.
- The public archive has no render-hardware or local-FPS evidence.
- The final render is on task but has overexposed sand, stray floating cubes, and detached-block polish issues.
- The final scene was reopened read-only but the final render was not independently rerendered or blind-evaluated.
Visible evidence gaps
- A strict one-user-turn rerun without recovery continuations
- A rerunnable executable model-authored Blender builder
- Render-hardware metadata and a local performance capture where relevant
- Independent rerender evidence and a 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.