Software engineering skill. Terminal-Bench 4.0 and SWE-bench Pro V2 HARD are the primaries. SWE-bench Verified is shown and ignored for the crown. As of Oct 14, 2026, Claude Opus 5.5 & Claude Sonnet 5.5 are tied for state of the art in Coding (since Oct 6, 2026; sources Terminal-Bench 4.0;…
| # | Model | State | Confidence |
|---|---|---|---|
| 1 | AClaude Opus 5.5 Anthropic | Co-SOTA · Coding | medium |
| 1 | AClaude Sonnet 5.5 Anthropic | Co-SOTA · Coding | medium |
| 2 | OGPT-6.1 Sol OpenAI | Near-SOTA · Coding | medium |
| 2 | OGPT-6 Astra OpenAI | Near-SOTA · Coding | medium |
| 3 | AClaude Fable 5.1 Anthropic | Near-SOTA · Coding | medium |
| 4 | AClaude Opus 5 Anthropic | #4 in Coding | medium |
| 5 | OGPT-6 Sol OpenAI | #5 in Coding | medium |
| 6 | AClaude Fable 5 Anthropic | #6 in Coding | medium |
| 7 | ZGLM 5.3 Zhipu AI | #7 in Coding | medium |
| 8 | xGrok 4.7 xAI | #8 in Coding | medium |
| 9 | OGPT-5.6 Sol OpenAI | #9 in Coding | medium |
| 10 | ZGLM 5.3 Flash Zhipu AI | #10 in Coding | medium |
| 11 | QQwen3 8 Max 0902 Alibaba / Qwen | #11 in Coding | medium |
| 12 | AClaude Opus 4.8 Anthropic | #12 in Coding | medium |
| 13 | OGPT-5.6 Terra OpenAI | #13 in Coding | medium |
Badges for the leaders
[](https://isitsota.com/m/claude-opus-5-5)
<a href="https://isitsota.com/m/claude-opus-5-5"><img alt="CO-SOTA · Coding · Oct 2026" src="https://isitsota.com/badge/claude-opus-5-5/coding.svg" height="28"></a>
[](https://isitsota.com/m/claude-sonnet-5-5)
<a href="https://isitsota.com/m/claude-sonnet-5-5"><img alt="CO-SOTA · Coding · Oct 2026" src="https://isitsota.com/badge/claude-sonnet-5-5/coding.svg" height="28"></a>
Not enough data yet
Gemini 4 Argon — #1 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Muse Wonder — #2 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Kimi K3 — #4 on SWE-bench Pro V2 HARD; missing Terminal-Bench 4.0 and LMArena Text (style-controlled, coding)
Muse Spark 1.2 — #9 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Claude Haiku 5.5 — #10 on SWE-bench Pro V2 HARD; missing Terminal-Bench 4.0 and LMArena Text (style-controlled, coding)
Claude Haiku 4.5 — #11 on SWE-bench Pro V2 HARD; missing Terminal-Bench 4.0 and LMArena Text (style-controlled, coding)
Mimo V2.6 Pro — #11 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Muse Spark 1.1 — #16 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Muse Spark — #18 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V4.1 Flash — #22 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Mimo V2.5 Pro — #25 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Gemini 3.1 Pro Preview — #26 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GPT-5.5 — #30 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Gemini 3.6 Flash — #32 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Gemini 3 Pro — #33 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Mimo V2.6 Flash — #37 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GPT-5.5 Instant — #39 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V4 Pro High — #47 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Kimi K2 5 Instant — #49 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Step 5 Preview — #50 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Mimo V2 Pro — #51 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Gemma 4 31B — #52 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Kimi K2 5 Thinking — #53 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Gemini 3.5 Flash Lite — #54 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Hy3 — #55 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 7 Plus — #58 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 8 27B — #59 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Mistral Large 4 — #60 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Minimax M3 — #63 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GPT-5.4 Mini — #64 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Inkling — #66 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Grok 4.1 — #67 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 5 397B A17b — #69 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Mimo V2.5 — #70 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V4 Pro High Preview — #72 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Grok 4.3 — #73 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Kimi K2 Thinking Turbo — #74 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GLM 4.7 — #75 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V4 Flash — #76 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Gemma 4 26B A4b — #77 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Minimax M2 7 — #79 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V4 Flash High Preview — #80 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Muse Glimmer — #81 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Mistral Medium 3.5 — #82 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V3.2 Thinking — #83 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V3.2 Exp Thinking — #84 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Nvidia Nemotron 3 Ultra 550B A55b Nvfp4 — #85 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 Max — #86 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Inkling Small — #87 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Longcat Flash Chat — #88 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 235B A22b Instruct 2507 — #89 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Mistral Large 3 — #90 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V3.2 — #92 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Kimi K2 0905 Preview — #93 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V3.2 Exp — #94 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek R1 0528 — #97 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 VL 235B A22b Instruct — #98 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GPT-5.4 Nano — #100 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Kimi K2 0711 Preview — #101 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V3.1 Terminus Thinking — #102 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 5 122B A10b — #103 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GPT-4.5 Preview — #104 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GLM 4.6 — #105 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 Coder 480B A35b Instruct — #106 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V3.1 Thinking — #108 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GLM 4.5 — #109 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 VL 235B A22b Thinking — #110 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 5 27B — #111 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Step 3.5 Flash — #112 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Claude Sonnet 4 — #113 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V3.1 — #114 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Mimo V2 Flash — #115 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Minimax M2 5 — #116 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 235B A22b NO Thinking — #117 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek R1 — #118 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 Next 80B A3b Instruct — #119 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Trinity Large Preview — #120 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 235B A22b Thinking 2507 — #121 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V3.1 Terminus — #122 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Minimax M2 1 Preview — #123 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 30B A3b Instruct 2507 — #124 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 5 Flash — #125 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Claude 3.5 Sonnet — #126 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Grok 4 0709 — #127 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 5 35B A3b — #128 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
O3 Mini — #129 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
O4 Mini — #130 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GPT-4.1 Mini — #131 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
O1 — #132 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Mistral Medium 2505 — #133 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 235B A22b — #134 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
DeepSeek V3 0324 — #137 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Gemini 2.5 Flash Preview 09 2025 — #138 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GLM 4.5 Air — #139 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Nvidia Nemotron 3.5 Lightning 30B A3b Nvfp4 — #140 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Gemini 2.5 Flash — #141 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GLM 4.7 Flash — #142 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Qwen3 Next 80B A3b Thinking — #143 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Minimax M1 — #144 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
GLM 4 6V — #145 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
O1 Preview — #146 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Intellect 3 — #147 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Trinity Large Thinking — #148 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD
Best page and recent changes
What can move this crown
- Terminal-Bench 4.0 — primary, weight 1
- SWE-bench Pro V2 HARD — primary, weight 1
- LMArena Text (style-controlled, coding) — primary, weight 0.6
- Terminal-Bench 2.0 — secondary, weight 0.35
- Aider Polyglot — secondary, weight 0.35
- SWE-bench Verified — display, weight 0 (display only, cannot crown)
Who is SOTA in Coding?
As of Oct 14, 2026, Claude Opus 5.5 & Claude Sonnet 5.5 are tied for state of the art in Coding (since Oct 6, 2026; sources Terminal-Bench 4.0; methodology v1.0).
Since when is the Coding crown current?
The current crown in Coding has held since Oct 6, 2026.