Space Flight Stage 2 — Landfall — GPT-6 Astra Max

Measured harness ledgerPublic result

Space Flight Stage 2 — Landfall — GPT-6 Astra Max

Extend the Stage 1 space-flight project with a seamless orbit-to-surface round trip, explorable terrain, takeoff, landing, and a fixed character asset using one frozen follow-on request.

Max reasoningHeadline result

Workflow cost: $188.63
Wall-clock: Not recorded
Processed tokens: 130.54M processed
Record state: partial_cumulative_stage1_and_landfall_root_metrics_clean_build_and_model_supplied_stage2_evidence

Public summary

GPT-6 Astra Max partial_cumulative_stage1_and_landfall_root_metrics_clean_build_and_model_supplied_stage2_evidence ledger: Not recorded wall-clock, 130.54M processed, and $188.63 Official OpenAI Standard API-list-price equivalent for the cumulative Stage 1 plus Landfall root-task receipt; not a Stage-2-only cost or an itemized Codex subscription charge.

Run identity and stack

Result ID: space-flight-game-threejs-stage-2-landfall-gpt-6-astra-max
Technical model: gpt-6-astra
Provider: OpenAI Codex
Stack: OpenAI Codex
Stack: Technical model/configuration: gpt-6-astra
Stack: Three.js Stage 2 workflow
Stack: Harness v1 Landfall prompt

Cost basis

Prompts above 272K input tokens use long-context pricing for the full request; all 1,101 supplied calls were short-context.

Recorded caveats

No valid wall-clock timing is available in the supplied receipt; no duration is reported here.
The published token counts and API-equivalent total are cumulative across Stage 1 and Landfall, not a Stage-2-only delta; subagent logs are excluded.
Output tokens include hidden reasoning, visible prose and code, and tool-call JSON.
Cache-read input is deeply discounted, so processed-token volume overstates cost.
The clean build verifies archive reproducibility, not gameplay behavior. The candidate's RC14 reports are model-supplied evidence, not independent verification.
The current source receipt does not byte-pin the provided Stage 1 baseline.

Visible evidence gaps

Stage-2-only token, timing, and all-agent measurement receipt
byte-pinned Stage 1 input baseline
independent real-input uninterrupted orbit-to-surface-to-orbit replay
independent repeatability replay
standardized FPS and transition-hitch receipt with hardware details
repair or explicitly accepted waiver for the close-ladder N camera-framing failure
surface-quality acceptance above the submitted minimum-3 failure
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.

Browse Builder Return to GPT-6 Astra