BuilderBuilder
GPT-6.1 Sol · Turbofan CAD
The episode says the Claude runs take about 44 minutes and about $13 and $14, Sol takes almost an hour and $3 to $6, and Astra costs the most. Those figures match the CAD-001 artifact metrics: Sonnet list-rate $13.28 and 44m 56s, Opus list-rate $14.07 and 44m 60s-scale 2699.6 seconds, Sol's API-equivalent range $3.24 to $6.14 and 57m 17s, and the nested unaided Astra estimate $31.78 and 57m 57s. model_finished is not recorded as a motion-contract pass. Supplied source capture Unaided model Open this round
Campaign evidence recordPublic result
GPT-6.1 Sol

Turbofan CAD — GPT-6.1 Sol

Create the CAD-001 v1.0.1 turbofan submission in the native CAD workflow.

Max reasoningHeadline result
Workflow cost
API-equivalent range $3.24–$6.14
Wall-clock
57m 17s
Processed tokens
Input 16,832,071, cached input 16,372,736, cache write 0, output 68,756. Cached input is not added to input. Reasoning output 34,602 is recorded separately and is not added. No processed total is recorded.
Record state
campaign_evidence_not_harness_ledger
Public summary

GPT-6.1 Sol Turbofan CAD: 57m 17s, API-equivalent range $3.24–$6.14, Input 16,832,071, cached input 16,372,736, cache write 0, output 68,756. Cached input is not added to input. Reasoning output 34,602 is recorded separately and is not added. No processed total is recorded..

Run identity and stack
  • Result ID: cad-001-gpt61-sol-max-2026-10-02
  • Technical model: gpt-6.1-sol
  • Provider: OpenAI Codex
  • Client: Codex CLI
  • Stack: CAD-001 turbofan fixture
  • Stack: Native CAD delivery
Cost basis
  • Clock is model_wall_clock_seconds 3437.246. Cost is api_equivalent_cost_range_usd [3.2435036, 6.1432272]. actual_subscription_charge_usd is null.
  • Source: usage.input_tokens, usage.cached_input_tokens, usage.cache_write_input_tokens, usage.output_tokens at harness commit 9b7c614dbb5103cc373fdf6275b5c4efc10d8f0e.
Validation evidence
  • Result: model_finished
  • Verification: Recorded status is model_finished. Pending, partial, stopped, qualified, and model-finished states stay unjudged. Campaign evidence. No top-level harness ledger. Sourced metrics stay attributed in cost, time, and the caveat. Clock is model_wall_clock_seconds 3437.246. Cost is api_equivalent_cost_range_usd [3.2435036, 6.1432272]. actual_subscription_charge_usd is null. Source: usage.input_tokens, usage.cached_input_tokens, usage.cache_write_input_tokens, usage.output_tokens at harness commit 9b7c614dbb5103cc373fdf6275b5c4efc10d8f0e.
Recorded caveats
  • model_finished on fixture cad-001-turbofan-v1.0.1. The cost is a range, not a point. Not a motion-contract pass. The older Sol plus Luna artifact is not this model.
  • Clock is model_wall_clock_seconds 3437.246. Cost is api_equivalent_cost_range_usd [3.2435036, 6.1432272]. actual_subscription_charge_usd is null.
  • Source: usage.input_tokens, usage.cached_input_tokens, usage.cache_write_input_tokens, usage.output_tokens at harness commit 9b7c614dbb5103cc373fdf6275b5c4efc10d8f0e.
Visible evidence gaps
  • campaign evidence; no top-level harness ledger
  • artifact-sourced metrics where a comparison record supplied them
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 BuilderReturn to GPT-6.1 Sol
RemakeBenchResearch console