AICoder
2026 Live Verified

AI Coding Models Leaderboard

Aggregated mirrors of 7 official benchmark boards and real-world coding task scenarios. 100% verified, objective, and zero paid placement.

7Boards
105Models ranked
4Evaluation Tasks
October 2026Period
2026-10-25Last update
●We fetch raw rankings from 5 public benchmark providers across 7 official boards — no re-weighting, no composite score. Click "Open official board" on each card to go to the source.
🧩Tool Matrix⚔️VS Compare💳Plans
ADSponsored

📌 7 official boards · mirrored 1:1

Each card previews a provider's Top 5. Click "View full" for the complete board, or jump to the official page. Data is 100% from each provider — we do no aggregation or math.

LMArena Code ★
Human preference blind battle coding Elo leaderboard
Coding
Metric:Coding Elo· 54 models
  1. #1Claude Opus 5.51813
  2. #2GPT-6 Astra1786.1
  3. #3Claude Sonnet 5.51774.3
  4. #4GPT-6.1 Sol1755.4
  5. #5Claude Fable 5.11743.7
Official ↗
LMArena Agent
Human preference on real agentic coding sessions
Agent
Metric:Agent Score· 37 models
  1. #1Claude Opus 5.515.53
  2. #2GPT-6 Astra13.09
  3. #3Claude Fable 5.112.66
  4. #4Claude Sonnet 5.511.95
  5. #5GPT-6.1 Sol11.72
Official ↗
CursorBench
Cursor's first-party agentic coding benchmark (v4.0)
Agent
Metric:Pass rate%· 17 models
  1. #1Claude Opus 5.557.8%
  2. #2Claude Sonnet 5.555.5%
  3. #3Gemini 4 Argon52.9%
  4. #4Claude Fable 5.151.8%
  5. #5Muse Wonder50.4%
Official ↗
Artificial Analysis
Independent industry standard Intelligence Index
Overall
Metric:Intelligence Index· 49 models
  1. #1Claude Opus 5.557.6
  2. #2Claude Sonnet 5.556
  3. #3Muse Wonder Full55.8
  4. #4Kimi K3.1 Pro55.1
  5. #5Claude Fable 5.153.4
Official ↗
Vals AI
Independent evaluations on real-world tasks in software, finance & cybersecurity
Overall
Metric:Accuracy%· 20 models
  1. #1Claude Sonnet 5.570.8%
  2. #2Claude Opus 5.570.2%
  3. #3Gemini 4 Argon69.2%
  4. #4Muse Wonder Full66.8%
  5. #5GPT-6 Astra65.0%
Official ↗
LiveBench
Contamination-resistant coding suite (completion, generation, JS, Python, TS)
Coding
Metric:Score%· 23 models
  1. #1GPT-6 Astra86.5%
  2. #2Claude Fable 5.186.1%
  3. #3Claude Opus 5.586.0%
  4. #4GPT-6.1 Sol85.6%
  5. #5Gemini 4 Argon81.5%
Official ↗
LMArena Text
Chatbot Arena text and general capability Elo rating
Chat
Metric:Arena Elo· 94 models
  1. #1Gemini 4 Argon1517
  2. #2Claude Opus 5.51509
  3. #3Muse Wonder1509
  4. #4Claude Fable 51503
  5. #5Claude Fable 5.11503
Official ↗

LMArena CodeCoding

Metric: Coding Elo·54 models·fetched 2026-10-12

Open official board ↗
✨

Not sure which model fits your budget & stack?

Try 4-Way Plan Showdowns & $20/mo vs API Cost Break-Even Meter

Compare Plans & Costs↗
#ModelCoding EloValue ScoreCross-board ranksActions
#1
Claude Opus 5.5Anthropic
1813
19.7
L #1
C #1
A #1
V #2
L #3
L #2
⚔️ VS💰Plans🚀Try
#2
GPT-6 AstraOpenAI
1786.1
7.9
L #2
C—
A #6
V #5
L #1
L #22
⚔️ VS💰Plans🚀Try
#3
Claude Sonnet 5.5Anthropic
1774.3
37.4
L #4
C #2
A #2
V #1
L #6
L #20
⚔️ VS💰Plans🚀Try
#4
GPT-6.1 SolOpenAI
1755.4
38.5
L #5
C—
A #9
V #7
L #4
L #19
⚔️ VS💰Plans🚀Try
#5
Claude Fable 5.1Anthropic
1743.7
7.6
L #3
C #4
A #5
V—
L #2
L #5
⚔️ VS💰Plans🚀Try
#6
Claude Opus 5Anthropic
1691.3
4.5
L #9
C #7
A—V—L—
L #9
⚔️ VS💰Plans🚀Try
#7
GPT-6 SolOpenAI
1688.2
38.3
L #6
C—A—V—L—
L #37
⚔️ VS💰Plans🚀Try
#8
Gemini 4 ArgonGoogle
1677.8
35.8
L #7
C #3
A #7
V #3
L #5
L #1
⚔️ VS💰Plans🚀Try
#9
Qwen3.8-MaxQwen
1673.5
43.8
L #21
C—
A #11
V #11
L #19
L #16
⚔️ VS💰Plans🚀Try
#10
Muse Spark 1.3Meta
1656.9
59.5
L #14
C #11
A #19
V #20
L #18
L #8
⚔️ VS💰Plans🚀Try
#11
Kimi K3Moonshot AI
1654.4
21.6
L #16
C—A—
V #18
L #15
L #11
⚔️ VS💰Plans🚀Try
#12
Grok 4.7xAI
1639.1
41.3
L #19
C #8
A #10
V #13
L #10
L #49
⚔️ VS💰Plans🚀Try
#13
Hunyuan-T1Tencent
1631.8
481.9
L #17
C—
A #29
V #16
L—
L #40
⚔️ VS💰Plans🚀Try
#14
DeepSeek-V4.1-FlashDeepSeek
1630
282.1
L—C—
A #12
V #9
L #7
L #23
⚔️ VS💰Plans🚀Try
#15
Claude Fable 5Anthropic
1624.7
7.5
L #8
C—A—V—L—
L #4
⚔️ VS💰Plans🚀Try
#16
GLM-5.3Zhipu AI
1622.2
56.5
L #22
C #9
A #13
V #10
L #17
L #18
⚔️ VS💰Plans🚀Try
#17
DeepSeek-V4-FlashDeepSeek
1618.6
406.7
L #15
C—A—V—L—
L #51
⚔️ VS💰Plans🚀Try
#18
GPT-5.6 SolOpenAI
1618.3
11.2
L #11
C #10
A—V—L—
L #14
⚔️ VS💰Plans🚀Try
#19
Grok 4.6xAI
1617.2
40.1
L #27
C #12
A—V—
L #12
L #26
⚔️ VS💰Plans🚀Try
#20
GLM-5.2Zhipu AI
1603.2
63.1
L #18
C—A—V—L—
L #24
⚔️ VS💰Plans🚀Try
#21
Qwen3.8-27BQwen
1592.8
256.6
L #30
C—
A #25
V—
L #21
L #52
⚔️ VS💰Plans🚀Try
#22
Gemini 3.7 FlashGoogle
1591.5
84.4
L #31
C—A—V—L—
L #10
⚔️ VS💰Plans🚀Try
#23
Claude Haiku 4.5Anthropic
1586.9
68
L—C—A—V—L—
L #61
⚔️ VS💰Plans🚀Try
#24
Gemini 3.8 FlashGoogle
1583.5
78
L #20
C #14
A #17
V #14
L #20
L #7
⚔️ VS💰Plans🚀Try
#25
DeepSeek-V4-ProDeepSeek
1582
120.5
L #24
C—
A #22
V #17
L—
L #33
⚔️ VS💰Plans🚀Try
#26
GPT-6 LunaOpenAI
1580.9
597.9
L #23
C—
A #21
V #19
L #16
L #46
⚔️ VS💰Plans🚀Try
#27
Claude Opus 4.8Anthropic
1557.5
14.2
L #10
C—A—V—L—
L #6
⚔️ VS💰Plans🚀Try
#28
Grok 4.5xAI
1552.8
42.3
L #25
C—A—V—L—
L #31
⚔️ VS💰Plans🚀Try
#29
Mistral Large 3Mistral AI
1541.3
35.2
L #33
C—
A #20
V—
L #23
L #53
⚔️ VS💰Plans🚀Try
#30
Claude Sonnet 5Anthropic
1540.8
28.6
L #12
C #16
A—V—L—
L #34
⚔️ VS💰Plans🚀Try
#31
Gemini 3.6 FlashGoogle
1537.8
74
L #35
C—A—V—L—
L #15
⚔️ VS💰Plans🚀Try
#32
Claude Sonnet 4.6Anthropic
1522.4
22.9
L—C—A—V—L—
L #28
⚔️ VS💰Plans🚀Try
#33
GPT-5.6 TerraOpenAI
1521.8
26.1
L #28
C #13
A #16
V—
L #13
L #29
⚔️ VS💰Plans🚀Try
#34
Doubao Seed 2.0 ProByteDance
1519.1
121.5
L—C—
A #36
V—L—
L #38
⚔️ VS💰Plans🚀Try
#35
GPT-5.6 LunaOpenAI
1519
232.4
L #29
C #15
A—V—L—
L #41
⚔️ VS💰Plans🚀Try
#36
Qwen3.7-MaxQwen
1515.1
198
L #32
C—
A #30
V—L—
L #27
⚔️ VS💰Plans🚀Try
#37
GPT-5.5OpenAI
1512.7
16.5
L #13
C—
A #26
V—L—
L #17
⚔️ VS💰Plans🚀Try
#38
Gemini 3.5 FlashGoogle
1510.5
40.5
L—C—A—V—L—
L #21
⚔️ VS💰Plans🚀Try
#39
Kimi K2.6Moonshot AI
1509.1
60.4
L—C—
A #44
V—L—
L #35
⚔️ VS💰Plans🚀Try
#40
GLM-5.1Zhipu AI
1508.3
62.5
L—C—A—V—L—
L #32
⚔️ VS💰Plans🚀Try
#41
Qwen3.6-PlusQwen
1482.3
202.1
L—C—
A #35
V—L—
L #36
⚔️ VS💰Plans🚀Try
#42
MiniMax-M3MiniMax
1481.7
211.8
L #34
C—
A #15
V #15
L #14
L #50
⚔️ VS💰Plans🚀Try
#43
Kimi K2.7 CodeMoonshot AI
1473.4
42.3
L—C—
A #28
V—L—L—
⚔️ VS💰Plans🚀Try
#44
GPT-5.4OpenAI
1465.3
27.6
L #26
C—A—V—L—
L #25
⚔️ VS💰Plans🚀Try
#45
Gemini 3.1 ProGoogle
1446.7
46.3
L #36
C—
A #27
V—L—
L #12
⚔️ VS💰Plans🚀Try
#46
Gemini 3.5 Flash-LiteGoogle
1439.7
124.8
L—C—
A #33
V—L—
L #39
⚔️ VS💰Plans🚀Try
#47
Gemini 3 ProGoogle
1439.5
49.3
L—C—A—V—L—
L #13
⚔️ VS💰Plans🚀Try
#48
GLM-4-PlusZhipu AI
1435
86.9
L—C—A—V—L—
L #48
⚔️ VS💰Plans🚀Try
#49
MiniMax-M2.7MiniMax
1397.9
215.1
L—C—A—V—L—
L #60
⚔️ VS💰Plans🚀Try
#50
GPT-5.4 miniOpenAI
1397.3
449.8
L—C—A—V—L—
L #42
⚔️ VS💰Plans🚀Try
#51
DeepSeek-V3DeepSeek
1362.2
630
L—C—A—V—L—
L #57
⚔️ VS💰Plans🚀Try
#52
Grok 4.3xAI
1357.1
71.9
L—C—A—V—L—
L #47
⚔️ VS💰Plans🚀Try
#53
Mistral Medium 3.5Mistral AI
1263.5
82.9
L #37
C—
A #37
V—L—
L #55
⚔️ VS💰Plans🚀Try
#54
Gemini 2.5 ProGoogle
1227.4
29.5
L—C—
A #23
V—L—
L #43
⚔️ VS💰Plans🚀Try
Rank color:#1#2#3#4-5#6-10#11+

The "Cross-board ranks" column shows this model's rank on the other boards at a glance (grey = not tested there). Boards use different methodologies, so raw values are not directly comparable.

LMArena Code board: LMSYS Coding Arena Elo ratings based on real developer blind tests on challenging code prompts. Sourced directly from arena.ai/leaderboard/code and mirrored 1:1.

ADSponsored
📊 7 benchmark boards
LMArena Code ↗ Coding

LMArena Code board: LMSYS Coding Arena Elo ratings based on real developer blind tests on challenging code prompts. Sourced directly from arena.ai/leaderboard/code and mirrored 1:1.

Coding Elo54 models
LMArena Agent ↗ Agent

LMArena Agent board: real agentic coding sessions scored by human blind votes (signal-based agent score). Sourced from arena.ai/leaderboard/agent and mirrored 1:1.

Agent Score37 models
CursorBench ↗ Agent

CursorBench is Cursor's first-party agentic coding benchmark (v4.0). Tasks come from real Cursor sessions with ambiguous, multi-file requirements. Metric: pass rate (%). Mirrored from Cursor's official cursor.com/cursorbench snapshot.

Pass rate%17 models
Artificial Analysis ↗ Overall

Artificial Analysis Intelligence Index: Tracks reasoning, quality, and coding intelligence across 600+ models, continuously updated as new frontier models release.

Intelligence Index49 models
Vals AI ↗ Overall

Vals AI (Vals Index): Independent benchmarks evaluating frontier AI models on economically valuable real-world tasks across software, finance, cybersecurity, and agent autonomy. Metric: Accuracy (%). Sourced directly from vals.ai and mirrored 1:1.

Accuracy%20 models
LiveBench ↗ Coding

LiveBench Coding: Monthly refreshed contamination-resistant suite extracting code completion, generation, JavaScript, Python, and TypeScript performance.

Score%23 models
LMArena Text ↗ Chat

LMArena Text board: The global standard LMSYS Chatbot Arena tracking cross-domain text generation and complex instruction-following Elo ratings from arena.ai/leaderboard/text.

Arena Elo94 models

All rankings and numbers come directly from official sources; we only do name matching and aggregate presentation, no weighting or re-ranking. Boards use different methodologies, so do not compare raw values across boards.