Who ranks #1 on the LLM Leaderboard leaderboard?
As of August 21, 2026, Gemini 3.7 Flash by Google ranks #1 on GPQA Diamond at 94.8%. API pricing is $0.75/M input and $3.75/M output.
LLM Leaderboard
🏆 LeaderboardAs of August 21, 2026, Gemini 3.7 Flash is #1 for the overall LLM ranking at 94.8%. Ranked by GPQA Diamond: graduate-level science questions written so Google cannot solve them. Live LLM leaderboard vs Chatbot Arena. Rank the best LLM by GPQA, HLE, SWE-bench, and Arena votes, with API pricing on the same row.
Updated August 21, 2026282 models33 providers
 Gemini 3.7 Flash Google · Proprietary | 94.8% | — | 97.2% | 80.8% | — | — | 71.2% | — | — | — | 1490 | $4.50 |
 Claude Mythos 5 Anthropic · Proprietary | 94.6% | 94.6% | — | 95.5% | — | — | — | — | — | — | — | $60.00 |
Gemini 3.1 Pro Google · Proprietary | 94.4% | 94.3% | 95.6% | 80.6% | 51.4% | — | 77.3% | — | — | — | 1486 | $17.50 |
 Kimi K3OSS Moonshot AI · Open Source | 93.5% | 93.5% | 68.9% | 93.4% | 56% | — | 42.7% | — | — | — | 1489 | $18.00 |
 Grok 4.5 xAI · Proprietary | 93.4% | 93% | 97.8% | 86.6% | — | — | 53.5% | — | — | — | — | $8.00 |
Grok 4.6 xAI · Proprietary | 93.2% | — | 99.2% | 95.6% | — | — | 54.4% | — | — | — | — | $8.00 |
Claude Opus 5 Anthropic · Proprietary | 92.9% | — | 93.3% | 96% | 64.7% | — | 56.7% | — | — | — | 1493 | $30.00 |
 Qwen3.8 MaxOSS Qwen · Open Source | 92.7% | 92.6% | 99.4% | 85.6% | 43.6% | — | 46.3% | — | — | — | 1491 | $8.00 |
 DeepSeek-V4-Pro-0813OSS DeepSeek · Open Source | 92.4% | — | — | 96.4% | 60% | — | — | — | — | — | — | $5.28 |
Gemini 3 Pro Google · Proprietary | 91.9% | 91.9% | 100% | 76.2% | 45.8% | — | 72.1% | — | 96.4% | — | 1485 | $14.00 |
 Muse Spark 1.1 Meta · Proprietary | 91.2% | — | — | 82% | 62.1% | — | — | — | — | — | 1489 | $5.50 |
DeepSeek-V4-Flash-0731OSS DeepSeek · Open Source | 91.0% | — | 94.4% | 88.8% | — | — | 34.7% | — | — | — | — | $1.76 |
Qwen3.7 Max Qwen · Proprietary | 90.9% | 92.4% | 95.6% | 80.4% | 41.4% | — | 58.5% | — | — | — | — | $5.00 |
 MiniMax M3OSS MiniMax · Open Source | 90.9% | — | 71.1% | 80.5% | — | — | — | — | — | — | — | $1.50 |
Kimi K2.6OSS Moonshot AI · Open Source | 90.8% | 90.5% | 96.1% | 80.2% | 36.4% | — | 38.7% | — | — | — | — | $4.93 |
 GLM-5.1OSS Z AI · Open Source | 89.9% | 86.2% | 93.3% | 74.2% | 52.3% | — | 37.3% | — | — | — | — | $5.80 |
DeepSeek-V4-Pro-MaxOSS DeepSeek · Open Source | 89.7% | 90.1% | 96.7% | 80.6% | 48.2% | — | 57.9% | — | — | — | — | $5.22 |
 GPT-5.4 OpenAI · Proprietary | 89.4% | 92.8% | 95.6% | 76.9% | 39.8% | — | 44.8% | — | — | — | — | $17.50 |
Gemini 3 Flash Google · Proprietary | 89.4% | 90.4% | 99.7% | 78% | 43.5% | — | 68.7% | — | — | — | — | $3.50 |
Gemini 3.5 Flash Google · Proprietary | 88.9% | — | 88.9% | 79.3% | 40.2% | — | 68.4% | — | — | — | 1477 | $10.50 |
Grok 4.3 xAI · Proprietary | 88.8% | — | 93.3% | 71.4% | — | — | 38% | — | — | — | — | $3.75 |
 Inkling-SmallOSS Thinking Machines · Open Source · via Thinking Machines Lab | 88.5% | 89.5% | 90% | 80.2% | 31.6% | — | 19.5% | — | — | — | — | $1.50 |
Qwen3.6 Plus Qwen · Proprietary | 88.4% | 90.4% | 93.3% | 78.8% | 28.8% | — | 49.1% | — | — | — | — | $3.50 |
Claude Opus 4.6 Anthropic · Proprietary | 88.4% | 91.3% | 99.8% | 80.8% | 53.1% | — | 41.0% | — | — | — | 1497 | $30.00 |
GPT-5.4 mini OpenAI · Proprietary | 88% | 88% | 88.9% | 73% | 28.2% | — | 28.6% | — | — | — | — | $5.25 |
Qwen3.7-Plus Qwen · Proprietary | 87.9% | 90.3% | 93.3% | 77.7% | 34.7% | — | — | — | — | — | — | $1.60 |
Kimi K2.7 CodeOSS Moonshot AI · Open Source | 87.9% | — | 95.6% | 78.2% | — | — | 39.2% | — | — | — | — | $4.93 |
GLM-5OSS Z AI · Open Source | 87.8% | — | 80% | 77.8% | — | — | — | — | — | — | — | $4.20 |
GPT-5.3 Codex OpenAI · Proprietary | 87.7% | 87.7% | — | 74.8% | — | — | — | — | — | — | — | $15.75 |
Grok-4 xAI · Proprietary | 87.5% | 87.5% | 91.7% | — | 40% | — | — | — | — | — | — | $18.00 |
MiniMax M2.7OSS MiniMax · Open Source | 86.6% | — | 91.0% | 73.8% | — | — | — | — | — | — | — | $1.50 |
Gemini 2.5 Pro Google · Proprietary | 86.4% | 83% | 83% | 63.2% | 17.8% | 88.9% | 50.8% | 88.2% | — | 90.9% | — | $11.25 |
Gemini 3.6 Flash Google · Proprietary | 86.4% | — | 82.2% | 79.6% | — | — | 68.7% | — | — | — | 1484 | $4.50 |
Claude Opus 4.7 Anthropic · Proprietary | 86.4% | 94.2% | 86.7% | 87.6% | 54.7% | — | 50.6% | — | — | — | 1494 | $30.00 |
GPT-5 OpenAI · Proprietary | 86.2% | 85.7% | 94.6% | 74.9% | 24.8% | 92.5% | 50.6% | 93.4% | 96% | — | — | $11.25 |
 Nemotron 3 Ultra (550B A55B)OSS NVIDIA · Open Source · via OpenRouter | 86.1% | 87% | — | 70.7% | 37.4% | — | — | — | — | — | — | $2.70 |
Qwen3.6-27BOSS Qwen · Open Source | 85.9% | 87.8% | 91.1% | 77.2% | 24% | — | — | — | — | — | — | $4.20 |
Qwen3.5-397B-A17BOSS Qwen · Open Source | 85.9% | 88.4% | 88.9% | 76.4% | 28.7% | 87.8% | — | — | — | — | — | $4.20 |
Claude Fable 5 Anthropic · Proprietary | 85.9% | — | 97.8% | 95% | 64.5% | — | 68.3% | — | — | — | 1506 | $60.00 |
Claude Opus 4.5 Anthropic · Proprietary | 85.5% | 87% | 48.1% | 80.9% | 25.2% | — | 41.8% | — | — | — | — | $30.00 |
Grok-4 Fast Reasoning xAI · Proprietary | 85.3% | — | 91.3% | 45.4% | — | — | — | — | — | — | — | $0.70 |
Claude Opus 4.8 Anthropic · Proprietary | 85.3% | 93.6% | 84.4% | 88.6% | 57.9% | — | 39.5% | — | — | — | 1481 | $30.00 |
Gemini 2.5 Pro Preview 06-05 Google · Proprietary | 84.8% | 86.4% | 88% | 67.2% | 21.6% | — | 54% | — | — | — | — | $11.25 |
Grok-3 xAI · Proprietary | 84.6% | 84.6% | 93.3% | — | — | 79.9% | — | — | 89.8% | — | — | $18.00 |
Grok-4.1 Fast Reasoning xAI · Proprietary | 84.3% | — | 91.9% | 41.4% | — | — | — | — | — | — | — | $0.70 |
Qwen3.6-35B-A3BOSS Qwen · Open Source · via OpenRouter | 83.8% | 86% | 86.7% | 73.4% | 21.4% | — | — | — | — | — | — | $1.14 |
GLM-4.7OSS Z AI · Open Source | 83.3% | 85.7% | 95.7% | 73.8% | 42.8% | 84.3% | 31.5% | 84.9% | — | 24.8% | — | $2.80 |
GPT-5.6 Sol OpenAI · Proprietary | 82.8% | 94.6% | 68.9% | 96.2% | — | — | 71.6% | — | — | — | 1481 | $35.00 |
GPT-5.4 nano OpenAI · Proprietary | 82.8% | 82.8% | 46.7% | 69.8% | 24.3% | — | 12% | — | — | — | — | $1.45 |
 MiMo-V2.5-ProOSS Xiaomi · Open Source | 82.6% | 66.7% | — | 78.9% | 34% | 89.4% | — | 75.6% | — | 89.8% | — | $1.30 |
Showing 1–50 of 282 models
Next step
You found the model. Now ship the product.
Auth, billing, and the API layer are already decided. Fork 8 finished AI apps, or have us build the first version with you.
The index
All Large Language Models
282 models across 33 providers. Search or jump to a lab — every model page stays linked here.
OpenAI
51 modelsGPT-5.6 CyberGPT-5.6 LunaGPT-5.6 SolGPT-5.6 TerraGPT-5.5 InstantGPT-5.5GPT-5.5 ProGPT-5.4 miniGPT-5.4 nanoGPT-5.4GPT-5.3 ChatGPT-5.3 CodexGPT-5.2 CodexGPT-5.2GPT-5.2 ProGPT-5.1 CodexGPT-5.1GPT-5.1 Codex HighGPT-5.1 Codex MiniGPT-5.1 HighGPT-5.1 InstantGPT-5.1 MediumGPT-5.1 ThinkingGPT-5 CodexGPT-5GPT-5 HighGPT-5 MediumGPT-5 miniGPT-5 nanoGPT OSS 120BGPT OSS 120B HighGPT OSS 20BGPT OSS 20B Higho3-proo3o4-miniGPT-4.1GPT-4.1 miniGPT-4.1 nanoo3-minio1o1-proo1-minio1-previewGPT-4oGPT-4o miniChatGPT-4o LatestGPT-4oGPT-4 TurboGPT-4GPT-3.5 Turbo
Anthropic
22 modelsClaude Opus 5Claude Sonnet 5Claude Fable 5Claude Mythos 5Claude Opus 4.8Claude Opus 4.7Claude Sonnet 4.6Claude Opus 4.6Claude Opus 4.5Claude Haiku 4.5Claude Sonnet 4.5Claude Opus 4.1Claude Opus 4Claude Sonnet 4Claude 3.7 SonnetClaude 3.5 HaikuClaude 3.5 SonnetClaude 3.5 SonnetClaude 3 HaikuClaude 3 OpusClaude 3 SonnetClaude Mythos Preview
Gemini 3.7 FlashGemini 3.5 Flash-LiteGemini 3.6 FlashGemini 3.5 FlashGemma 4 26B-A4BGemma 4 31BGemini 3.1 Flash-LiteGemini 3.1 ProGemini 3 FlashGemini 3 ProGemma 3n E4B InstructedGemini 2.5 Flash-LiteGemini 2.5 Pro Preview 06-05Gemini 2.5 FlashGemini 2.5 ProGemma 3 12BGemma 3 27BGemma 3 4BGemini 2.0 Flash-LiteGemini 2.0 FlashGemma 2 27BGemini 1.5 FlashGemini 1.5 ProGemini 1.5 Flash 8BGemini 1.0 Pro
Cursor
2 modelsxAI
15 modelsDeepSeek
16 modelsQwen
38 modelsQwen3.8-27BQwen3.8-2.4T-A95BQwen3.8 MaxQwen3.7 FlashQwen3.7-PlusQwen3.7 MaxQwen3.6-27BQwen3.6-35B-A3BQwen3.6 PlusQwen3.5-9BQwen3.5-122B-A10BQwen3.5-27BQwen3.5-35B-A3BQwen3.5-397B-A17BQwen3 MaxQwen3 VL 235B A22B InstructQwen3 VL 235B A22B ThinkingQwen3 VL 30B A3B InstructQwen3 VL 30B A3B ThinkingQwen3 VL 32B InstructQwen3 VL 4B InstructQwen3 VL 4B ThinkingQwen3 VL 8B InstructQwen3 VL 8B ThinkingQwen3-Next-80B-A3B-InstructQwen3-Next-80B-A3B-ThinkingQwen3-235B-A22B-Thinking-2507Qwen3-235B-A22B-Instruct-2507Qwen3 235B A22BQwen3 30B A3BQwen3 32BQwen3-Coder 480B A35B InstructQwen2.5 VL 72B InstructQwen3-CoderQwQ-32B-PreviewQwen2.5 72B InstructQwen2.5 7B InstructQwen2.5-Coder 32B Instruct
AI21 Labs
2 models
Amazon
5 models
Baidu
1 models
ByteDance
2 models
Cohere
3 models
IBM
1 models
Inception
1 models
inclusionAI
1 models
LG AI Research
1 models
Meituan
4 modelsMeta
12 models
Mistral
23 modelsMistral Medium 3.5Mistral Small 4Min istral 3 (3B Reasoning 2512)MiniStral 3 (14B Instruct 2512)Ministral 3 (14B Reasoning 2512)Ministral 3 (3B Instruct 2512)Ministral 3 (8B Instruct 2512)Ministral 3 (8B Reasoning 2512)Mistral Large 3 (675B Instruct 2512)Mistral Large 3Devstral Small 1.1Devstral MediumMistral Small 3.2 24B InstructMagistral MediumMistral Small 3.1 24B BaseMistral Small 3.1 24B InstructMistral Small 3 24B InstructPixtral LargeMinistral 8B InstructMistral SmallPixtral-12BMistral Large 2Mistral NeMo Instruct
Moonshot AI
9 models
Nous Research
1 modelsNVIDIA
5 models
Poolside
2 models
Sakana AI
1 models
StepFun
1 models
Tencent
1 modelsThinking Machines
1 models
Unisound
1 models
Upstage
1 modelsXiaomi
3 modelsWhat is the best LLM?
As of August 21, 2026, Gemini 3.7 Flash by Google is #1 for the overall LLM ranking at 94.8%. Ranked by GPQA Diamond: graduate-level science questions written so Google cannot solve them. This board also tracks GPQA Diamond, GPQA, AIME 2025, SWE-bench Verified. Next on the same board: Claude Mythos 5 and Gemini 3.1 Pro. Related leaders: GPT-5.6 Sol on GPQA at 94.6%; GPT-5.2 on AIME 2025 at 100%. This llm leaderboard leaderboard ranks models by GPQA Diamond. Scores come from public evals. Prices are the live API rates in the table above.
Sources: GPQA Diamond (GitHub) (https://github.com/idavidrein/gpqa); OpenAI API pricing (https://developers.openai.com/api/docs/pricing); Anthropic Claude API pricing (https://platform.claude.com/docs/en/about-claude/pricing); Gemini API pricing (https://ai.google.dev/gemini-api/docs/pricing)
| Rank | Model | GPQA Diamond | Input /M | Output /M |
|---|---|---|---|---|
| 1 | Gemini 3.7 Flash | 94.8% | $0.75 | $3.75 |
| 2 | Claude Mythos 5 | 94.6% | $10.00 | $50.00 |
| 3 | Gemini 3.1 Pro | 94.4% | $2.50 | $15.00 |
| 4 | Kimi K3 | 93.5% | $3.00 | $15.00 |
| 5 | Grok 4.5 | 93.4% | $2.00 | $6.00 |
| 6 | Grok 4.6 | 93.2% | $2.00 | $6.00 |
| 7 | Claude Opus 5 | 92.9% | $5.00 | $25.00 |
| 8 | Qwen3.8 Max | 92.7% | $2.00 | $6.00 |
What is Chatbot Arena?
Chatbot Arena is the LMSYS / LMArena human-vote leaderboard. Two models answer the same prompt. Voters pick a winner without seeing the names. That Elo is useful for chat quality and a poor proxy for SWE-bench or computer use. This page keeps Arena Elo next to exam scores and API price.
Claude vs ChatGPT: which LLM is smarter?
Smarter on what. Claude vs ChatGPT (GPT) trades the lead across evals. Use this overall board for a general rank, the coding board for repo work, and the pricing table for token cost. Brand loyalty is a bad routing rule.
Related LLM leaderboards
Individual benchmark pages
Rank one eval at a time. All LLM benchmarks.
LLM Leaderboard FAQ
What is the best LLM?
As of August 21, 2026, Gemini 3.7 Flash by Google is #1 for the overall LLM ranking at 94.8%. Ranked by GPQA Diamond: graduate-level science questions written so Google cannot solve them. This board also tracks GPQA Diamond, GPQA, AIME 2025, SWE-bench Verified. Next on the same board: Claude Mythos 5 and Gemini 3.1 Pro. Related leaders: GPT-5.6 Sol on GPQA at 94.6%; GPT-5.2 on AIME 2025 at 100%.
What are the top models on GPQA Diamond?
The current GPQA Diamond ranking as of August 21, 2026 is 1. Gemini 3.7 Flash at 94.8%; 2. Claude Mythos 5 at 94.6%; 3. Gemini 3.1 Pro at 94.4%.
Which llm leaderboard model is the cheapest?
Llama 3.1 8B Instruct is the cheapest scored model on this llm leaderboard leaderboard at $0.03/M input and $0.03/M output ($0.06 blended). Gemini 3.7 Flash still leads GPQA Diamond at 94.8%.
Should I always pick the #1 GPQA Diamond model?
Not automatically. Gemini 3.7 Flash leads GPQA Diamond, but a cheaper scored model can be the better production choice if the quality gap is small. Use the table to weigh GPQA Diamond against input/output price, context window, and related evals.
How often is the LLM Leaderboard leaderboard updated?
Scores and API prices on this page are refreshed from published evals and provider rates. The snapshot is labeled August 21, 2026. Treat it as a current index, not a one-off blog post.
What is the best LLM overall?
On this LLM leaderboard as of August 21, 2026, Gemini 3.7 Flash ranks #1 on GPQA Diamond at 94.8%. “Best” still depends on the job: use the coding board for SWE-bench, the pricing table for token cost, and this overall board for a general ranking.
What is an LLM leaderboard?
An LLM leaderboard ranks language models on shared benchmarks such as GPQA, MMLU, Humanity's Last Exam, and SWE-bench, then shows API price next to each score so you can compare quality and cost.
How is this LLM leaderboard different from LMSYS Chatbot Arena?
Chatbot Arena ranks models from blind human votes. This page ranks the same frontier models on public evals and published token prices, including Arena Elo when we have it.
Which LLM benchmark should I sort by?
Use GPQA or HLE for reasoning, SWE-bench for coding, and Arena Elo for human preference. Sort the table to switch the ranking without leaving the page.
What is the best LLM right now?
There is no single best LLM. Rank this overall board for a general score, then open Coding for SWE-bench, Agentic for computer use, and Pricing when token cost is the constraint. Chatbot Arena is a separate human-vote rank.
What is Chatbot Arena?
Chatbot Arena (LMSYS / LMArena) ranks models from blind human votes, not from exams. Arena Elo is the preference score. This page also shows public evals and API price on the same row, which Arena does not.
What is the LMSYS leaderboard?
The LMSYS leaderboard is Chatbot Arena. Humans pick a winner between two hidden models. Use it for chat feel. Use GPQA, HLE, or SWE-bench when you need a task score, and this table when you also need $/M.
How is this LLM leaderboard different from Chatbot Arena?
Arena is votes. This LLM leaderboard is published evals plus live API rates. We include Arena Elo when we have it, then keep SWE-bench, HLE, and GPQA next to input and output price so you can pick a production SKU.
Claude vs ChatGPT: which LLM is better?
Claude vs ChatGPT is a job split. Rank coding on SWE-bench, agents on OSWorld, writing on Arena Elo, then compare token price. ChatGPT Plus is a chat subscription. Claude API and GPT API are token bills.