BuilderBuilder
Measured harness ledgerPublic result
Qwen3.8 Max

Infinite cathedral corridor shader — Qwen3.8 Max xhigh

Generate an infinite stained-glass cathedral corridor shader with convincing depth, architectural repetition, light, and motion.

xhigh reasoningHeadline result
Workflow cost
¥82.08 Independently recomputed first-party API-list-price equivalent for the measured token mix; NOT a billed amount. This run used an Alibaba Model Studio token-plan subscription, under which usage is not itemized as cash. (upper bound)
Wall-clock
43m 30.8s wall-clock
Processed tokens
6.63M processed
Record state
artifact_independent_boot_ledger_incomplete_shader_metrics
Public summary

Qwen3.8 Max xhigh artifact_independent_boot_ledger_incomplete_shader_metrics ledger: 43m 30.8s wall-clock, 6.63M processed, and ¥82.08 Independently recomputed first-party API-list-price equivalent for the measured token mix; NOT a billed amount. This run used an Alibaba Model Studio token-plan subscription, under which usage is not itemized as cash. (upper bound).

Run identity and stack
  • Result ID: infinite-cathedral-shadertoy-qwen3.8-max-xhigh
  • Technical model: qwen3.8-max
  • Provider: Alibaba Cloud Model Studio token plan
  • Client: Qwen Code 0.21.6 installed / 0.21.6 transcript-recorded
  • Stack: Alibaba Cloud Model Studio token plan
  • Stack: Qwen Code 0.21.6 installed / 0.21.6 transcript-recorded
  • Stack: Technical model/configuration: qwen3.8-max
  • Stack: ShaderToy fragment shader
  • Stack: Harness v1 infinite cathedral prompt
Cost basis
  • This row headlines the full-input-rate UPPER BOUND, matching the Stage 1 STARFALL and Courtyard Arena rows. The Stage 2 Landfall and JRPG rows headline the 10% cache-hit rule instead. Convert before comparing any two Qwen rows.
  • Qwen Code records no per-request cost field; null by construction.
  • Computed here for comparability: applying the pricing page's generic 10% cache-hit rule (¥1.2/M) to the 6,292,330 cache-read tokens gives ¥14.12. The receipt itself reports only the ¥82.08 upper bound.
Primary artifact integrity
  • Kind: manifest-verified-artifact
  • Path: artifacts/infinite-cathedral-shadertoy-qwen3.8-max-xhigh/source/cathedral.glsl
  • SHA-256: ed7fc2506fc1a4aa51537ea0d2594025dd82fd625c06717811e5de0ed6ce7dcb
Recorded caveats
  • Wall-clock (2,610.8 s) is end-to-end workflow latency including tool execution and browser checks — not model-only compute. Summed API duration (2,649.0 s) is separate and exceeds it because subagents ran concurrently.
  • One Qwen Code agent turn fans out to many billable requests and subagent sessions under a single session id (43 root + 12 subagent here).
  • Visible output and reasoning are separate fields; both are output-priced on this stack.
  • inputTokens includes cached tokens and outputTokens includes reasoning tokens; the ledger's totalTokens and total_processed are different quantities and must never be swapped.
  • Cache reads are normally discounted; pricing them at the fresh-input rate makes ¥82.08 an upper bound.
  • Qwen Code records no cache-write counter, so that line item is unavailable rather than zero.
  • Context occupancy is a single-request footprint, not cumulative usage; quota fields are null with no explicit before/after evidence.
  • The client version is reported as-is: installed and transcript-recorded both read 0.21.6 while the marker expected 0.21.7.
  • Four of 53 tool calls errored during the run; all 55 model requests returned HTTP 200.
  • The task manifest requires an initial/optimized local FPS pair and an RTX Pro 6000 1080p60 final render. Neither exists for this result — see task_requirements_status. This row is not a complete shader measurement.
  • The single 46.2 fps figure is Apple M1 Pro / ANGLE Metal evidence at 1920x993, not a standardized cross-machine comparison.
  • The shader was verified in a local WebGL2 harness, not on shadertoy.com itself, so hosted-Shadertoy compilation is unconfirmed.
  • The sanitized session export named in receipt.sha256 is withheld from the public archive by its own privacy label.
Visible evidence gaps
  • initial and optimized local FPS recorded under identical conditions
  • the RTX Pro 6000 1080p60 final render
  • local frame target and final-render environment fields
  • confirmation that the shader compiles on shadertoy.com itself
  • 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 BuilderReturn to Qwen3.8 Max
RemakeBenchResearch console