LMArena.ai — Top Models July 2026
LMArena.ai is the rebranded LMSys Chatbot Arena. Same blind pairwise-voting methodology, same Elo math, new home. Here is who leads each board this month and what the rebrand actually changed for buyers.
The LMSys to LMArena.ai story
The Chatbot Arena began in 2023 as a research project under LMSys, an academic group out of UC Berkeley. It quickly became the most-cited LLM benchmark because it measured something the capability-only benchmarks could not: actual human preference under blind side-by-side comparison. By 2024 the project had processed millions of votes, become a procurement input for Fortune 500 buyers, and outgrown its original academic scaffold. The 2024-25 transition to the lmarena.ai domain consolidated the project as an independent organisation while keeping the same Elo methodology and open vote pool.
For users the rebrand changed almost nothing: same prompts, same blind voting, same Elo math. For procurement teams the rebrand codified Arena Elo as a vendor-neutral signal independent of any single university. That is what made it sticky as a reference point in enterprise contracts.
The four-way race at the top
The July 2026 snapshot below shows Claude Opus 4.8 as the new text and coding leader, with Gemini 3.1 Pro and Opus 4.7 just behind at the historical 1500 Elo barrier. The top of LMArena.ai is now genuinely contested.
Top of LMArena.ai text leaderboard (July 2026) Claude Opus 4.8 Thinking 1512 ████████████████████ text & coding #1 Gemini 3.1 Pro Preview 1500 ███████████████████ science leader Claude Opus 4.7 Thinking 1495 ███████████████████ prior coding #1 GPT-5.5 Pro 1488 ██████████████████ reasoning DeepSeek V4 Pro 1462 █████████████████ Apache 2.0 Qwen 3.7 Max 1455 ████████████████ top Chinese model Claude Sonnet 4 1402 ██████████████ workhorse tier GPT-4.1 1395 █████████████ legacy frontier Gemini 2.5 Pro 1388 █████████████ legacy frontier Llama 4 Maverick 1352 ███████████ open weights Mistral Large 3 1341 ███████████ open weights
Full Leaderboard
| # | Model | Quality | Arena ELO | Speed | Price | Context | Value | Released |
|---|---|---|---|---|---|---|---|---|
| 1 | Anthropic · Frontier agentic coding & knowledge work | 100 | 1525 | 58 t/s | $10 / $50 | 1M | 3.3 | Jun 2026 |
| 2 | Anthropic · Coding, agents & computer use | 99 | 1512 | 72 t/s | $5 / $25 | 1M | 6.6 | May 2026 |
| 3 | OpenAI · Reasoning at any cost | 98 | 1510 | 68 t/s | $30 / $180 | 1M | 0.9 | Apr 2026 |
| 4 | OpenAI: GPT-5.6 New OpenAI · General frontier reasoning & tools | 98 | 1514 | 96 t/s | $5 / $30 | 400K | 5.6 | Jul 2026 |
| 5 | OpenAI · Frontier general purpose | 97 | 1506 | 70 t/s | $5 / $30 | 1M | 5.5 | Apr 2026 |
| 6 | OpenAI · Complex analysis | 97 | — | — | $30 / $180 | 1M | 0.9 | Mar 2026 |
| 7 | OpenAI · Complex analysis | 97 | — | — | $21 / $168 | 400K | 1.0 | Dec 2025 |
| 8 | Anthropic · Complex analysis | 97 | — | — | $30 / $150 | 1M | 1.1 | May 2026 |
| 9 | Google · Long-context & multimodal | 97 | 1508 | 128 t/s | $2 / $12 | 2M | 13.9 | Jul 2026 |
| 10 | Anthropic · Coding & agentic workflows | 96 | 1505 | 68 t/s | $5 / $25 | 1M | 6.4 | Apr 2026 |
| 11 | OpenAI · Deep research | 96 | — | — | $10 / $40 | 200K | 3.8 | Oct 2025 |
| 12 | OpenAI · Deep research | 96 | — | — | $2 / $8 | 200K | 19.2 | Oct 2025 |
| 13 | OpenAI · Hard reasoning | 96 | — | — | $20 / $80 | 200K | 1.9 | Jun 2025 |
| 14 | Google · Speed & cost | 96 | 1505 | — | $2 / $12 | 1M | 13.7 | Feb 2026 |
| 15 | Google · Science & long-context | 96 | 1505 | 131 t/s | $2 / $12 | 1M | 13.7 | Apr 2026 |
| 16 | Anthropic · General purpose | 95 | 1490 | — | $5 / $25 | 1M | 6.3 | Feb 2026 |
| 17 | Anthropic · General purpose | 95 | — | — | $5 / $25 | 200K | 6.3 | Nov 2025 |
| 18 | Anthropic · Complex analysis | 95 | — | — | $30 / $150 | 1M | 1.1 | Apr 2026 |
| 19 | Google · Image generation | 94 | — | — | $2 / $12 | 66K | 13.4 | Nov 2025 |
| 20 | Anthropic · Multimodal | 94 | — | — | $15 / $75 | 200K | 2.1 | Aug 2025 |
| 21 | OpenAI · Hard reasoning | 94 | 1370 | 68 t/s | $10 / $40 | 200K | 3.8 | Apr 2025 |
| 22 | Alibaba Cloud · Long autonomous agentic runs | 94 | 1488 | 90 t/s | $2.5 / $7.5 | 1M | 18.8 | May 2026 |
| 23 | xAI · Agentic tasks & real-time info | 93 | 1496 | 83 t/s | $1.25 / $2.5 | 1M | 49.6 | May 2026 |
| 24 | OpenAI · General purpose | 93 | 1495 | — | $2.5 / $15 | 1M | 10.6 | Mar 2026 |
| 25 | OpenAI · General purpose | 93 | — | — | $1.75 / $14 | 128K | 11.8 | Mar 2026 |
| 26 | OpenAI · Code generation | 93 | — | — | $1.75 / $14 | 400K | 11.8 | Feb 2026 |
| 27 | OpenAI · Code generation | 93 | — | — | $1.75 / $14 | 400K | 11.8 | Jan 2026 |
| 28 | OpenAI · General purpose | 93 | — | — | $1.75 / $14 | 128K | 11.8 | Dec 2025 |
| 29 | OpenAI · General purpose | 93 | — | — | $1.75 / $14 | 400K | 11.8 | Dec 2025 |
| 30 | OpenAI · Code generation | 93 | — | — | $1.25 / $10 | 400K | 16.5 | Dec 2025 |
| 31 | OpenAI · General purpose | 93 | — | — | $1.25 / $10 | 400K | 16.5 | Nov 2025 |
| 32 | OpenAI · General purpose | 93 | — | — | $1.25 / $10 | 128K | 16.5 | Nov 2025 |
| 33 | OpenAI · Code generation | 93 | — | — | $1.25 / $10 | 400K | 16.5 | Nov 2025 |
| 34 | OpenAI · Hard reasoning | 93 | — | — | $150 / $600 | 200K | 0.2 | Mar 2025 |
| 35 | OpenAI · Complex analysis | 93 | — | — | $30 / $60 | 8K | 2.1 | May 2023 |
| 36 | OpenAI · Multimodal | 93 | — | — | $30 / $60 | 8K | 2.1 | May 2023 |
| 37 | xAI · General purpose | 93 | 1496 | — | $1.25 / $2.5 | 2M | 49.6 | Mar 2026 |
| 38 | OpenAI · Complex analysis | 93 | — | — | $8 / $15 | 272K | 8.1 | Apr 2026 |
| 39 | Anthropic · Balanced agents & coding value | 93 | 1479 | 98 t/s | $3 / $15 | 1M | 10.3 | Jun 2026 |
| 40 | Moonshot AI · Frontier quality at low cost | 92 | 1466 | 48 t/s | $0.73 / $3.49 | 256K | 43.6 | Apr 2026 |
| 41 | Google · Multimodal + value | 92 | 1345 | 87 t/s | $1.25 / $10 | 1M | 16.4 | Mar 2025 |
| 42 | DeepSeek · Open-weight reasoning value | 92 | 1471 | 62 t/s | $0.5 / $1.1 | 256K | 115.0 | Jul 2026 |
| 43 | Anthropic · Complex analysis | 91 | 1360 | 52 t/s | $15 / $75 | 200K | 2.0 | May 2025 |
| 44 | · Hard reasoning | 91 | — | — | $0.3 / $1.1 | 164K | 130.0 | Jul 2025 |
| 45 | Google · Speed & cost | 91 | — | — | $1.25 / $10 | 1M | 16.2 | Jun 2025 |
| 46 | DeepSeek · Hard reasoning | 91 | — | — | $0.5 / $2.15 | 164K | 68.7 | May 2025 |
| 47 | Google · Speed & cost | 91 | — | — | $1.25 / $10 | 1M | 16.2 | May 2025 |
| 48 | DeepSeek · Hard reasoning | 91 | — | — | $0.29 / $0.29 | 33K | 313.8 | Jan 2025 |
| 49 | DeepSeek · Hard reasoning | 91 | — | — | $0.7 / $0.8 | 131K | 121.3 | Jan 2025 |
| 50 | DeepSeek: R1OSS DeepSeek · Hard reasoning | 91 | — | — | $0.7 / $2.5 | 64K | 56.9 | Jan 2025 |
| 51 | Moonshot AI · Open-weight agentic coding | 91 | — | 55 t/s | $0.73 / $3.49 | 256K | 43.1 | Jun 2026 |
| 52 | · Open-weight reasoning & tool use | 91 | — | 50 t/s | $0.2 / $0.8 | 262K | 182.0 | Jun 2026 |
| 53 | Meta · Open-weight general & multilingual | 91 | 1466 | 76 t/s | $0.8 / $2.4 | 1M | 56.9 | Jun 2026 |
| 54 | DeepSeek · Open-source value leader | 90 | 1467 | 33 t/s | $1.74 / $3.48 | 1M | 34.5 | Apr 2026 |
| 55 | Anthropic · Coding & balance | 90 | 1467 | 73 t/s | $3 / $15 | 1M | 10.0 | Feb 2026 |
| 56 | OpenAI · General purpose | 90 | 1455 | — | $1.25 / $10 | 400K | 16.0 | Aug 2025 |
| 57 | xAI · General purpose | 90 | — | — | $3 / $15 | 131K | 10.0 | Apr 2025 |
| 58 | Alibaba Cloud · Open-source | 90 | — | — | $1.04 / $6.24 | 262K | 24.7 | Apr 2026 |
| 59 | OpenAI · Long context | 89 | 1310 | 120 t/s | $2 / $8 | 1M | 17.8 | Apr 2025 |
| 60 | Moonshot AI · Speed & cost | 89 | 1452 | — | $0.4 / $1.9 | 262K | 77.4 | Jan 2026 |
| 61 | MiniMax · Open-weight agentic coding | 89 | 1455 | 80 t/s | $0.6 / $2.4 | 1M | 59.3 | Jun 2026 |
| 62 | · Open-weight agentic coding (provisional) | 89 | — | — | $0.98 / $3.08 | 200K | 43.8 | Jun 2026 |
| 63 | · Open-weight agentic & tool use | 88 | 1467 | 48 t/s | $0.98 / $3.08 | 200K | 43.3 | Apr 2026 |
| 64 | OpenAI · Multimodal | 88 | — | — | $10 / $10 | 400K | 8.8 | Oct 2025 |
| 65 | OpenAI · Complex analysis | 88 | — | — | $15 / $120 | 400K | 1.3 | Oct 2025 |
| 66 | Anthropic · General purpose | 88 | — | — | $3 / $15 | 1M | 9.8 | Sep 2025 |
| 67 | OpenAI · General purpose | 88 | — | — | $2.5 / $10 | 128K | 14.1 | Aug 2025 |
| 68 | OpenAI · Search + citations | 88 | — | — | $2.5 / $10 | 128K | 14.1 | Mar 2025 |
| 69 | OpenAI · Hard reasoning | 88 | — | — | $15 / $60 | 200K | 2.3 | Dec 2024 |
| 70 | OpenAI · General purpose | 88 | — | — | $2.5 / $10 | 128K | 14.1 | Nov 2024 |
| 71 | OpenAI · General purpose | 88 | — | — | $2.5 / $10 | 128K | 14.1 | May 2024 |
| 72 | OpenAI · Multimodal | 88 | — | — | $6 / $18 | 128K | 7.3 | May 2024 |
| 73 | OpenAI · General purpose | 88 | — | — | $5 / $15 | 128K | 8.8 | May 2024 |
| 74 | OpenAI · Multimodal | 88 | — | — | $10 / $30 | 128K | 4.4 | Apr 2024 |
| 75 | OpenAI · Complex analysis | 88 | — | — | $10 / $30 | 128K | 4.4 | Jan 2024 |
| 76 | OpenAI · Multimodal | 88 | — | — | $10 / $30 | 128K | 4.4 | Nov 2023 |
| 77 | Z.ai: GLM 5OSS · Open-source | 88 | 1450 | — | $0.6 / $1.92 | 80K | 69.8 | Feb 2026 |
| 78 | Anthropic · Coding & balance | 88 | 1320 | 95 t/s | $3 / $15 | 200K | 9.8 | May 2025 |
| 79 | OpenAI · Reasoning & math | 88 | 1305 | 155 t/s | $1.1 / $4.4 | 200K | 32.0 | Jan 2025 |
| 80 | xAI · Real-time info | 87 | 1330 | 82 t/s | $3 / $15 | 131K | 9.7 | Feb 2025 |
| 81 | DeepSeek · Open-source | 87 | 1455 | — | $0.252 / $0.378 | 164K | 276.2 | Dec 2025 |
| 82 | · Open-source | 86 | — | — | $0.135 / $0.5 | 131K | 270.9 | Dec 2025 |
| 83 | DeepSeek · Open-source | 86 | — | — | $0.287 / $0.431 | 164K | 239.6 | Dec 2025 |
| 84 | DeepSeek · Open-source | 86 | — | — | $0.27 / $0.41 | 164K | 252.9 | Sep 2025 |
| 85 | DeepSeek · Open-source | 86 | — | — | $0.27 / $0.95 | 164K | 141.0 | Sep 2025 |
| 86 | DeepSeek · Open-source | 86 | — | — | $0.21 / $0.79 | 33K | 172.0 | Aug 2025 |
| 87 | DeepSeek · Open-source | 86 | — | — | $0.2 / $0.77 | 164K | 177.3 | Mar 2025 |
| 88 | Anthropic · General purpose | 86 | — | — | $3 / $15 | 200K | 9.6 | Feb 2025 |
| 89 | Anthropic · Hard reasoning | 86 | — | — | $3 / $15 | 200K | 9.6 | Feb 2025 |
| 90 | DeepSeek · Best open-source value | 86 | 1310 | 62 t/s | $0.27 / $1.1 | 128K | 125.5 | Mar 2025 |
| 91 | Alibaba Cloud · Multilingual & APAC | 86 | 1448 | 124 t/s | $1.4 / $5.6 | 256K | 24.6 | Apr 2026 |
| 92 | OpenAI · General purpose | 85 | 1285 | 109 t/s | $2.5 / $10 | 128K | 13.6 | May 2024 |
| 93 | Mistral AI · Open-source | 85 | — | — | $0.5 / $1.5 | 262K | 85.0 | Dec 2025 |
| 94 | Mistral AI · Open-source | 85 | — | — | $2 / $6 | 131K | 21.3 | Nov 2024 |
| 95 | Mistral AI · Open-source | 85 | — | — | $2 / $6 | 128K | 21.3 | Feb 2024 |
| 96 | Google · Speed & cost | 84 | — | — | $1.5 / $9 | 1M | 16.0 | May 2026 |
| 97 | · Accessible open-weight agentics | 84 | — | 110 t/s | $0.05 / $0.2 | 262K | 672.0 | Jun 2026 |
| 98 | OpenAI · Speed & cost | 83 | — | — | $0.75 / $4.5 | 400K | 31.6 | Mar 2026 |
| 99 | OpenAI · Speed & cost | 83 | — | — | $0.25 / $2 | 400K | 73.8 | Aug 2025 |
| 100 | Alibaba Cloud · Open-source | 82 | — | — | $0.04 / $0.15 | 256K | 863.2 | Mar 2026 |
What to do this quarter
- Update bookmarks and citations. Internal eval-spec docs and procurement RFPs that reference "lmsys.org" should be updated to lmarena.ai. The data continues at the new domain.
- Pull from the right board. Coding teams should cite the coding Arena Elo (Claude Opus 4.8 now leads at ~1582, ahead of Opus 4.7 at 1567). Generic chat teams should cite the text leaderboard (Opus 4.8 leads at ~1512, Gemini 3.1 Pro Preview at ~1500).
- Build dual-vendor capability. The top four models are within 40 Elo of each other. Treat them as interchangeable on capability and optimise for switching cost.
- Pair Arena scores with workload-specific evals. Arena rewards short-conversation polish. Long-context, tool-use, and domain-specific tasks need their own measurement.
- Track the open-weight gap. DeepSeek V4 Pro under Apache 2.0 sits at 1462 Elo, within 38 points of the text leader. The gap is the smallest it has ever been.
- Watch GPT-5.5 Pro pricing. At $30/$180 per 1M tokens, paying for the top of LMArena.ai now costs 200x more per token than the cheapest tier. The cost curve is steepening.
- Re-baseline at every model launch. Tokenizer changes between releases (as seen across the Claude Opus 4.6 → 4.7 → 4.8 line) shift effective cost without shifting list price.
Related reading
- LMArena Explained — what LMArena is, how to read Arena Elo
- AI Model Leaderboard — full quality, speed, pricing comparison
- LLM Leaderboard
- LMSys Arena Leaderboard May 2026
- LMArena Elo Explained for Enterprise Buyers
Teams running side-by-side evals against multiple LMArena.ai leaders typically expose them through Swfte Connect as a single endpoint, then run their own internal Elo on production prompts. That is the only way to verify whether public Arena rank translates to your workload.