In 2026, Claude Opus 4.8 tops our intelligence index at 55.7, just ahead of OpenAI’s GPT-5.5 (54.8). Google’s Gemini 3.5 Flash (50.2) and open-weight GLM-5.2 (51.1) sit close behind, with Gemini 3.1 Pro (46.5) and DeepSeek V4 Pro (44.3) rounding out the frontier. The single “best” model depends on how you weigh intelligence against price and speed.
The Intelligence column is a composite 0-100 score based on the Artificial Analysis Intelligence Index v4.1, which blends nine demanding evaluations spanning reasoning, coding, agentic tool use, and scientific knowledge. Higher is smarter — but a two-point gap at the very top rarely changes real-world output quality, so treat clusters of models as roughly equivalent rather than reading tiny differences as decisive.
Read intelligence alongside price and speed. A frontier model like Opus 4.8 costs far more per token than DeepSeek V4 Flash (40.3) or GLM-5.2, which deliver 70-90% of the intelligence at a fraction of the cost. For high-volume or latency-sensitive work, a cheaper, faster model usually wins; for the hardest reasoning and agentic tasks, the top of the table earns its premium. Sort by the metric that matches your use case.
| # | النموذج ↕ | المطوّر | الذكاء ↕ | السياق ↕ | التكلفة لكل مليون رمز (مدخلات) ↕ | التكلفة لكل مليون رمز (مخرجات) ↕ | Open weights |
|---|---|---|---|---|---|---|---|
| 1 | Claude Opus 5 | أنثروبيك | 61 | مليون | $5.00 | $25.00 | لا |
| 2 | Claude Fable 5 | أنثروبيك | 60 | مليون | $10.00 | $50.00 | لا |
| 3 | GPT-5.6 Sol | OpenAI | 59 | 1.05 مليون رمز | $5.00 | $30.00 | لا |
| 4 | Kimi K3 | مون شوت آي | 57 | مليون | $3.00 | $15.00 | نعم |
| 5 | Claude Opus 4.8 | أنثروبيك | 55.7 | مليون | $5.00 | $25.00 | لا |
| 6 | GPT-5.5 | OpenAI | 54.8 | 1.05 مليون رمز | $5.00 | $30.00 | لا |
| 7 | GLM 5.2 | زهي بو آي | 51.1 | مليون | $1.40 | $4.40 | نعم |
| 8 | Gemini 3.5 Flash | جوجل | 50.2 | مليون | $1.50 | $9.00 | لا |
| 9 | Claude Sonnet 4.6 | أنثروبيك | 47 | مليون | $3.00 | $15.00 | لا |
| 10 | Gemini 3.1 Pro | جوجل | 46.5 | 1.05 مليون رمز | $2.00 | $12.00 | لا |
| 11 | DeepSeek V4-Pro | DeepSeek | 44.3 | مليون | $0.44 | $0.87 | نعم |
| 12 | Kimi K2.7 Code | مون شوت آي | 42 | 256 ألف رمز | $0.60 | $2.50 | نعم |
| 13 | DeepSeek V4-Flash | DeepSeek | 40.3 | مليون | $0.14 | $0.28 | نعم |
| 14 | Claude Haiku 4.5 | أنثروبيك | 37 | 200 ألف | $1.00 | $5.00 | لا |
| 15 | DeepSeek R1 | DeepSeek | 20.1 | 128 ألف رمز | $0.50 | $2.15 | نعم |
| 16 | Mistral Large 3 | ميسترال إيه آي | 15.9 | 256 ألف رمز | $2.00 | $6.00 | نعم |
| 17 | Llama 4 Maverick | ميتا | 14.3 | مليون | $0.15 | $0.60 | نعم |
| 18 | Qwen3 235B-A22B | علي بابا | 13 | 128 ألف رمز | $0.45 | $1.80 | نعم |
| 19 | Qwen3 32B | علي بابا | 12 | 128 ألف رمز | $0.08 | $0.28 | نعم |
| 20 | Llama 4 Scout | ميتا | 10.0 | 10 ملايين | $0.10 | $0.30 | نعم |
| 21 | Gemma 3 27B | جوجل | 7.4 | 128 ألف رمز | $0.08 | $0.16 | نعم |
| 22 | Phi-4 | مايكروسوفت | 4.9 | 16 ألف رمز | $0.07 | $0.14 | نعم |
| 23 | Claude Sonnet 5 | أنثروبيك | — | مليون | $2.00 | $10.00 | لا |
| 24 | DeepSeek R1 Distill Llama 70B | DeepSeek | — | 128 ألف رمز | $0.80 | $0.80 | نعم |
| 25 | Gemini 2.5 Pro | جوجل (جوجل ديب مايند) | — | مليون رمز (1,048,576 رمزًا) | $1.25 | $10.00 | لا |
| 26 | Gemini 3.6 Flash | جوجل | — | مليون | $1.50 | $7.50 | لا |
| 27 | Gemma 3 12B | جوجل | — | 128 ألف رمز | $0.05 | $0.15 | نعم |
| 28 | Gemma 3 4B | جوجل | — | 128 ألف رمز | $0.05 | $0.10 | نعم |
| 29 | Grok 4 | إكس إيه آي | — | 256,000 رمز | $3.00 | $15.00 | لا |
| 30 | Kling 2.5 Turbo Pro | Kuaishou | — | — | — | — | لا |
| 31 | Llama 3.1 8B | ميتا | — | 128 ألف رمز | $0.02 | $0.03 | نعم |
| 32 | Llama 3.3 70B | ميتا | — | 128 ألف رمز | $0.10 | $0.32 | نعم |
| 33 | Mistral 7B | ميسترال إيه آي | — | 32K | $0.02 | $0.03 | نعم |
| 34 | Mistral NeMo 12B | ميسترال إيه آي | — | 128 ألف رمز | $0.02 | $0.04 | نعم |
| 35 | NVIDIA Nemotron 3 Nano Omni | NVIDIA | — | 256 ألف رمز | — | — | نعم |
| 36 | Qwen3 14B | علي بابا | — | 128 ألف رمز | $0.12 | $0.24 | نعم |
| 37 | Qwen3 30B-A3B | علي بابا | — | 128 ألف رمز | $0.12 | $0.50 | نعم |
| 38 | Qwen3 8B | علي بابا | — | 128 ألف رمز | $0.04 | $0.14 | نعم |
| 39 | Sora 2 | OpenAI | — | — | — | — | لا |
| 40 | Sora 2 Pro | OpenAI | — | — | — | — | لا |
| 41 | Veo 3.1 | جوجل | — | — | — | — | لا |
| 42 | Wan 2.5 | علي بابا | — | — | — | — | لا |
Click any column heading to re-sort. Intelligence = a 0–100 composite of public reasoning/knowledge benchmarks; — means not yet scored. Prices are USD per 1M tokens. Updated August 2026.
الأسئلة الشائعة
What is the best LLM right now?
As of mid-2026, Claude Opus 4.8 is the highest-scoring model on the Artificial Analysis Intelligence Index (55.7), narrowly ahead of GPT-5.5 (54.8). Both lead on reasoning, coding, and agentic tasks, so the practical choice between them usually comes down to price, speed, and ecosystem rather than a meaningful intelligence gap.
What is the smartest open-source model?
GLM-5.2 is currently the most intelligent open-weight model, scoring 51.1 on the intelligence index — ahead of DeepSeek V4 Pro (44.3) and DeepSeek V4 Flash (40.3). That puts the best open models within roughly four points of proprietary frontier models like GPT-5.5, while remaining free to self-host or run through low-cost APIs.
Which AI model is the best value?
DeepSeek V4 Flash and GLM-5.2 offer the strongest intelligence-per-dollar. DeepSeek V4 Flash scores 40.3 at a tiny fraction of Opus 4.8’s price, delivering roughly 70% of the top score for a small percentage of the cost. For budget-sensitive, high-volume workloads, these open models are the clear value leaders.
How is the intelligence score calculated?
Our scores mirror the Artificial Analysis Intelligence Index v4.1, a 0-100 composite of nine demanding evaluations — including Humanity’s Last Exam, GPQA Diamond, Terminal-Bench, SciCode, and agentic tool-use tasks. It measures reasoning, knowledge, coding, and agentic ability in one number, so a higher score means broadly more capable across hard, real-world tasks.
