Ranking by author picks
21| # | Model | Coding posts | Author picks | Median stated cost (n) | Median stated time (n) | Most tested in | Top matchups |
|---|
| 1 | Claude Opus 5.5Anthropic | 191 | 60 of 71 (85%) | $10.17 (85) | 47 min (69) | 3D & Shaders, Simulations | vs Gemini 4 Argon
vs GPT-6 Astra
vs DeepSeek V4.1 Flash 1017 |
|---|
| 2 | DeepSeek V4.1 Flash 1017DeepSeek | 101 | 14 of 33 (42%) | $0.46 (32) | 57.5 min (36) | 3D & Shaders, Animations/SVG | vs Claude Opus 5.5
vs Gemini 4 Argon
vs GPT-6.1 Sol |
|---|
| 3 | Claude Sonnet 5.5Anthropic | 32 | 6 of 9 (67%) | $7.36 (9) | 1 h 03 min (9) | 3D & Shaders, Simulations | vs Claude Opus 5.5
vs GPT-6.1 Sol
vs GPT-6 Astra |
|---|
| 4 | GPT-6.1 SolOpenAI | 56 | 6 of 20 (30%) | $1.13 (27) | 21 min (29) | 3D & Shaders, Simulations | vs Claude Opus 5.5
vs Gemini 4 Argon
vs DeepSeek V4.1 Flash 1017 |
|---|
| 5 | GPT-6 AstraOpenAI | 75 | 6 of 30 (20%) | $13.17 (26) | 23 min (19) | 3D & Shaders, Simulations | vs Claude Opus 5.5
vs Gemini 4 Argon
vs Claude Sonnet 5.5 |
|---|
| 6 | Gemini 4 ArgonGoogle | 99 | 4 of 34 (12%) | $1.50 (37) | 24 min (34) | 3D & Shaders, Animations/SVG | vs Claude Opus 5.5
vs GPT-6 Astra
vs GPT-6.1 Sol |
|---|
| 7 | Kimi K3Moonshot AI | 16 | 2 of 5 (40%) | $15.58 (3) | — | 3D & Shaders, Games | vs Claude Opus 5.5
vs GPT-6 Astra
vs Gemini 4 Argon |
|---|
| 8 | Claude Opus 5Anthropic | 12 | 1 of 1 (100%) | $10.00 (3) | 2 h (4) | Games, 3D & Shaders | vs GPT-5.6 Sol
vs DeepSeek V4.1 Flash 1017
vs Kimi K3 |
|---|
| 9 | GPT-6 LunaOpenAI | 7 | 1 of 1 (100%) | — | 43 s (3) | 3D & Shaders, Animations/SVG | vs Claude Haiku 5.5
vs GPT-6 Sol
vs GPT-6 Astra |
|---|
| 10 | Gemini 4 ProGoogle | 6 | 1 of 1 (100%) | — | — | 3D & Shaders, Simulations | vs GPT-6 Astra
vs Claude Opus 5.5 |
|---|
| 11 | GLM 5.3Zhipu AI (Z.ai) | 5 | 1 of 1 (100%) | — | — | Animations/SVG, 3D & Shaders | vs Claude Opus 5.5 |
|---|
| 12 | GLM 5.3 FlashZhipu AI (Z.ai) | 12 | 1 of 2 (50%) | — | 27 min (3) | 3D & Shaders, Web/UI | vs DeepSeek V4.1 Flash 1017
vs Gemini 3.8 Flash
vs Claude Haiku 5.5 |
|---|
| 13 | GPT-6 SolOpenAI | 9 | 1 of 2 (50%) | — | — | 3D & Shaders, Games | vs GPT-6 Luna
vs DeepSeek V4.1 Flash 1017
vs GPT-6 Astra |
|---|
| 14 | DeepSeek V4 Pro 0813DeepSeek | 6 | 1 of 2 (50%) | — | — | Games, 3D & Shaders | vs Claude Opus 5.5
vs GPT-6.1 Sol |
|---|
| 15 | Claude Fable 5Anthropic | 13 | 1 of 3 (33%) | $2.16 (4) | 20 min (3) | Games, 3D & Shaders | vs GPT-5.6 Sol
vs Kimi K3
vs Claude Opus 4.8 |
|---|
| 16 | DeepSeek V4.1 FlashDeepSeek | 24 | 0 of 11 (0%) | $0.12 (8) | 30.9 min (6) | 3D & Shaders, Animations/SVG | vs DeepSeek V4.1 Flash 1017
vs Claude Opus 5.5
vs GPT-6 Astra |
|---|
| 17 | Claude Haiku 5.5Anthropic | 17 | 0 of 4 (0%) | $0.43 (5) | 10 min (5) | 3D & Shaders, Animations/SVG | vs Claude Opus 5.5
vs GPT-6 Luna
vs DeepSeek V4.1 Flash 1017 |
|---|
| 18 | Grok 4.7xAI | 17 | 0 of 5 (0%) | $3.98 (5) | 1 h 16 min (10) | 3D & Shaders, Simulations | vs Claude Opus 5.5
vs Gemini 4 Argon
vs GPT-6.1 Sol |
|---|
| 19 | Muse Wonder 1.0Meta | 14 | 0 of 5 (0%) | $1.64 (9) | 12 min (9) | 3D & Shaders, Games | vs Gemini 4 Argon
vs Claude Opus 5.5
vs GPT-6.1 Sol |
|---|
| 20 | GPT-5.6 SolOpenAI | 10 | 0 of 1 (0%) | $1.53 (3) | — | 3D & Shaders, Animations/SVG | vs Claude Fable 5
vs Kimi K3
vs Claude Opus 5 |
|---|
| 21 | Claude Fable 5.1Anthropic | 8 | 0 of 3 (0%) | — | — | Animations/SVG, 3D & Shaders | vs GPT-6 Astra
vs Claude Opus 5.5 |
|---|
Author picks: posts with two or more different models where the author named a winner; “12 of 20” means the model was picked in 12 of the 20 such posts it appears in. Medians use the cost or time each post states for that model (shown with at least 3 values); tasks, prompts and settings differ, so they are not normalised.
Best AI for each kind of coding
- Games: Claude Opus 5.5 (4), Gemini 4 Argon (2), GPT-6.1 Sol (2) — picks across 14 of 62 posts that name a winner.
- Web/UI: Claude Opus 4.1 (1), Claude Opus 5.5 (1), Kimi K3 (1) — picks across 3 of 22 posts that name a winner.
- 3D & Shaders: Claude Opus 5.5 (36), DeepSeek V4.1 Flash 1017 (6), Claude Sonnet 5.5 (3) — picks across 60 of 202 posts that name a winner.
- Simulations: Claude Opus 5.5 (12), Claude Sonnet 5.5 (2), DeepSeek V4.1 Flash 1017 (2) — picks across 20 of 44 posts that name a winner.
- Animations/SVG: Claude Opus 5.5 (7), DeepSeek V4.1 Flash 1017 (5), Claude Fable 5.5 (2) — picks across 17 of 50 posts that name a winner.
How this ranking works
Every row is computed from the posts in the archive: a model is credited when the post’s author says its result was the better one. Posts where every model “won”, solo runs and harness tests that ran one model on both sides are left out of the tally. Read the posts themselves before you choose — thematchup pages show every side-by-side video, and the model pages list every post per model. Updated with each collection run; last data change 2026-10-19.