Human Score
Points this season. Higher is better.
beckon run gauntlet --season 1 --tests 8
The vibe coder's benchmark
Eight one-shot tests, identical conditions. Every prompt, artifact, and score public.
Runs live inside BeckonToken efficiency
Output tokens per point. Lower is better.
Time on the bench
Wall clock across scored tests. Lower is better.
Leaderboard human, canonical
Scored on camera: runs first try, polish, prompt adherence, wow factor. Ten points per test. Ties go to the cheaper run.
| Model | 01 | 02 | 03 | 04 | 05 | 06 | 07 | 08 | Total | |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | gpt-5.6 (Sol Ultra) OpenAI |
8 | 10 | · | 10 | · | 10 | · | 10 | 48/50 |
| 2 | claude-fable-5 (Max effort) Anthropic |
4 | 6 | · | 4 | · | 10 | · | 10 | 34/50 |
◌ run captured, scoring pending. Chips open the full test result.
Head-to-heads watch the runs
Same prompt, same clock, side by side.
01 · Haunted Horror Game
gpt-5.6 Sol Ultra 8 vs claude-fable-5 Max effort 4Full result
AI panel claude-fable-5 Max effort 2–1 — the panel overturned the human verdictVotes
“It delivers denser atmosphere via a full torch-and-darkness system, battery management, multi-state AI hunter, richer Web Audio (drones, whispers, presence, stings), and stronger visual polish with grain, vignettes, and scare/win sequences.” — grok-4-5
02 · Lava Lamp
gpt-5.6 Sol Ultra 10 vs claude-fable-5 Max effort 6Full result
AI panel gpt-5.6 Sol Ultra 3–0Votes
04 · SVG Self-Portrait
gpt-5.6 Sol Ultra 10 vs claude-fable-5 Max effort 4Full result
AI panel claude-fable-5 Max effort 3–0 — the panel overturned the human verdictVotes
“It perfectly captures the essence of a language model with its brilliant conceptual design, using text paths and typography to represent the flow of prompts and tokens.” — gemini-3-1-pro
06 · Remotion Promo Video
gpt-5.6 Sol Ultra 10 vs claude-fable-5 Max effort 10Full result
AI panel gpt-5.6 Sol Ultra 2–1Votes
08 · The Blender Build (live)
gpt-5.6 Sol Ultra 10 vs claude-fable-5 Max effort 10Full result
AI panel claude-fable-5 Max effort 2–1Votes
Efficiency recorded, never scored
| Model | Points | Time | Output tokens | Tokens / point | Cost |
|---|---|---|---|---|---|
| gpt-5.6 (Sol Ultra) | 48 | 72m 33s | 107k | 2.2k | · |
| claude-fable-5 (Max effort) | 34 | 125m 9s | 1186k | 34.9k | · |
Arena AI judge panel
| Model | ELO | Record | |
|---|---|---|---|
| 1 | claude-fable-5-max | 1012 | 3W–2L |
| 2 | gpt-5-6-sol-ultra | 988 | 2W–3L |
5 matches. Every vote is public on the Arena page.