Best AI Models for Product Managers
Frontier models ranked by the PM Index — our product-management-weighted blend of Artificial Analysis benchmarks — next to price and context window. Use the slider to re-rank by quality or value for your budget.
Updated June 26, 2026
Model recommender
Which AI model should I use?
Pick your task and budget for a data-backed recommendation.
Recommended for writing prds
Claude Opus 4.8
See the full ranking for Writing PRDs →PM Index 55.9 · $10.00/1M tok
Value frontier chart
Higher = smarter (PM Index), left = cheaper. Ringed models are the best value — nothing beats them on both. Hover a point, or a lab below.
- Anthropic
- OpenAI
- Z AI
- Alibaba
- DeepSeek
- MiniMax
- Kimi
- Xiaomi
Movers · last 7 days
▲ Climbing
- Solar Open 100B▲2 · #68
- Mistral Medium 3.1▲2 · #69
- Magistral Medium 1.2▲1 · #63
- Nova 2.0 Pro Preview▲1 · #64
- Kimi K2▲1 · #65
▼ Falling
- NVIDIA Nemotron 3 Nano 30B A3B▼8 · #76
- Nova 2.0 Lite▼4 · #67
- GLM-5-Turbo▼1 · #29
- Kimi K2.5▼1 · #30
- Gemini 3 Flash Preview▼1 · #31
★ New
AI model rankings
Best qualityBest value
| # | Model | PM Index | AA Intelligence | $/1M tok | Value |
|---|---|---|---|---|---|
| 1 | Muse Spark Meta | 41.8 | 43.1 | — | 10.0 |
| 2 | DeepSeek V4 Pro DeepSeek | 44.2 | 44.3 | $0.54 | 9.9 |
| 3 | MiniMax-M3 MiniMax | 44.0 | 44.4 | $0.52 | 9.9 |
| 4 | DeepSeek V4 Flash DeepSeek | 40.1 | 40.3 | $0.18 | 9.8 |
| 5 | MiMo-V2.5 Xiaomi | 40.1 | 40.1 | $0.18 | 9.8 |
| 6 | MiMo-V2.5-Pro Xiaomi | 41.4 | 42.2 | $0.54 | 9.7 |
| 7 | GLM-5-Turbo Z AI | 38.1 | 38.1 | — | 9.7 |
| 8 | GLM-5.2 Z AI | 51.3 | 51.1 | $2.15 | 9.6 |
| 9 | MiniMax-M2.7 MiniMax | 37.1 | 38.1 | $0.52 | 9.3 |
| 10 | Qwen3.7 Plus Alibaba | 37.1 | 39.0 | $0.59 | 9.3 |
| 11 | Hy3-preview Tencent | 33.6 | 33.6 | $0.20 | 9.2 |
| 12 | MiMo-V2-Flash Xiaomi | 33.2 | 33.2 | $0.15 | 9.2 |
| 13 | Qwen3.6 Plus Alibaba | 38.7 | 39.6 | $1.13 | 9.1 |
| 14 | DeepSeek V3.2 DeepSeek | 33.4 | 33.4 | $0.34 | 9.1 |
| 15 | MiMo-V2-Pro Xiaomi | 40.3 | 40.3 | $1.50 | 9.0 |
| 16 | Kimi K2.6 Kimi | 41.6 | 42.8 | $1.71 | 9.0 |
| 17 | Grok Build 0.1 0616 xAI | 38.6 | 39.8 | $1.25 | 9.0 |
| 18 | Gemini 3 Flash Preview Google | 37.8 | 37.8 | $1.13 | 9.0 |
| 19 | Kimi K2.5 Kimi | 38.1 | 38.1 | $1.19 | 9.0 |
| 20 | MiniMax-M2.5 MiniMax | 33.7 | 33.7 | $0.52 | 9.0 |
| 21 | GLM-5 Z AI | 39.5 | 39.5 | $1.55 | 8.9 |
| 22 | Nemotron 3 Ultra 550B A55B NVIDIA | 36.9 | 37.8 | $1.18 | 8.9 |
| 23 | Command A+ Cohere | 29.3 | 29.3 | $0.00 | 8.9 |
| 24 | GPT-5.4 mini OpenAI | 39.7 | 40.0 | $1.69 | 8.9 |
| 25 | Qwen3.6 27B Alibaba | 36.7 | 37.1 | $1.35 | 8.8 |
| 26 | MiniMax-M2.1 MiniMax | 31.4 | 31.4 | $0.52 | 8.8 |
| 27 | Gemini 3.5 Flash Google | 49.6 | 50.2 | $3.38 | 8.8 |
| 28 | GLM-4.7 Z AI | 33.8 | 33.8 | $1.00 | 8.7 |
| 29 | Step 3.7 Flash StepFun | 28.8 | 29.7 | $0.44 | 8.6 |
| 30 | GLM-5.1 Z AI | 39.7 | 40.2 | $2.15 | 8.6 |
| 31 | Kimi K2 Thinking Kimi | 32.7 | 32.7 | $1.07 | 8.6 |
| 32 | Grok 4.3 xAI | 35.4 | 37.6 | $1.56 | 8.5 |
| 33 | Step 3.5 Flash 2603 StepFun | 26.0 | 26.0 | $0.15 | 8.5 |
| 34 | MiniMax-M2 MiniMax | 28.3 | 28.3 | $0.52 | 8.5 |
| 35 | Step 3.5 Flash StepFun | 25.5 | 25.5 | $0.15 | 8.5 |
| 36 | K-EXAONE LG AI Research | 24.7 | 24.7 | — | 8.4 |
| 37 | DeepSeek V3.2 Exp DeepSeek | 25.4 | 25.4 | $0.31 | 8.4 |
| 38 | EXAONE 4.5 33B LG AI Research | 23.0 | 23.0 | — | 8.3 |
| 39 | ERNIE 5.0 Thinking Preview Baidu | 21.9 | 21.9 | — | 8.2 |
| 40 | Qwen3.6 Max Preview Alibaba | 40.0 | 40.0 | $2.92 | 8.2 |
| 41 | Qwen3.7 Max Alibaba | 44.9 | 46.0 | $3.75 | 8.1 |
| 42 | Nemotron Cascade 2 30B A3B NVIDIA | 21.3 | 21.3 | — | 8.1 |
| 43 | NVIDIA Nemotron 3 Super 120B A12B NVIDIA | 23.3 | 25.4 | $0.41 | 8.1 |
| 44 | Mistral Small 4 Mistral | 20.8 | 20.8 | $0.26 | 8.0 |
| 45 | Grok 4.20 0309 v2 xAI | 37.0 | 37.0 | $3.00 | 7.8 |
| 46 | Sonar Reasoning Pro Perplexity | 17.8 | 17.8 | — | 7.8 |
| 47 | Grok 4.20 0309 xAI | 36.5 | 36.5 | $3.00 | 7.8 |
| 48 | GPT-5.1 OpenAI | 38.9 | 38.9 | $3.44 | 7.8 |
| 49 | Kimi K2 0905 Kimi | 23.5 | 23.5 | $1.07 | 7.7 |
| 50 | GPT-5.4 OpenAI | 51.3 | 51.4 | $5.63 | 7.6 |
| 51 | Gemini 3.1 Pro Preview Google | 43.7 | 46.5 | $4.50 | 7.6 |
| 52 | Solar Open 100B Upstage | 15.1 | 15.1 | — | 7.6 |
| 53 | DeepSeek V3.1 Terminus DeepSeek | 26.3 | 26.3 | $1.92 | 7.5 |
| 54 | Kimi K2 Kimi | 19.4 | 19.4 | $1.03 | 7.4 |
| 55 | Solar Pro 2 (Preview) Upstage | 12.5 | 12.5 | — | 7.3 |
| 56 | Nova 2.0 Lite Amazon | 17.3 | 20.5 | $0.85 | 7.3 |
| 57 | Solar Pro 3 Upstage | 12.0 | 14.1 | — | 7.3 |
| 58 | GPT-5.2 OpenAI | 42.2 | 42.2 | $4.81 | 7.3 |
| 59 | Sonar Reasoning Perplexity | 11.7 | 11.7 | — | 7.2 |
| 60 | Llama Nemotron Super 49B v1.5 NVIDIA | 12.4 | 12.4 | $0.18 | 7.2 |
| 61 | NVIDIA Nemotron 3 Nano 30B A3B NVIDIA | 11.8 | 14.2 | $0.10 | 7.2 |
| 62 | Gemini 3 Pro Preview Google | 39.6 | 39.6 | $4.50 | 7.2 |
| 63 | EXAONE 4.0 32B LG AI Research | 10.6 | 10.6 | — | 7.1 |
| 64 | Mistral Medium 3.5 Mistral | 29.4 | 29.9 | $3.00 | 7.1 |
| 65 | Mistral Medium 3.1 Mistral | 14.8 | 14.8 | $0.80 | 7.1 |
| 66 | Mistral Large 3 Mistral | 14.2 | 15.9 | $0.75 | 7.1 |
| 67 | Claude Sonnet 4.6 Anthropic | 47.5 | 47.2 | $6.00 | 7.0 |
| 68 | Sonar Perplexity | 9.5 | 9.5 | — | 7.0 |
| 69 | Sonar Pro Perplexity | 9.3 | 9.3 | — | 7.0 |
| 70 | Llama 4 Maverick Meta | 11.9 | 14.3 | $0.47 | 7.0 |
| 71 | Solar Pro 2 Upstage | 9.0 | 9.0 | — | 7.0 |
| 72 | Granite 4.1 30B IBM | 7.5 | 8.9 | — | 6.9 |
| 73 | Granite 4.1 8B IBM | 7.1 | 6.7 | $0.06 | 6.8 |
| 74 | Nova Lite Amazon | 6.9 | 6.9 | $0.10 | 6.8 |
| 75 | Llama 4 Scout Meta | 8.0 | 10.0 | $0.29 | 6.8 |
| 76 | R1 1776 Perplexity | 6.3 | 6.3 | — | 6.8 |
| 77 | ERNIE 4.5 300B A47B Baidu | 9.0 | 9.0 | $0.49 | 6.8 |
| 78 | Solar Mini Upstage | 6.2 | 6.2 | $0.15 | 6.7 |
| 79 | Gemini 2.5 Pro Google | 27.0 | 27.0 | $3.44 | 6.7 |
| 80 | Llama 3.3 Instruct 70B Meta | 8.6 | 8.6 | $0.61 | 6.6 |
| 81 | Granite 4.0 H Small IBM | 5.2 | 5.2 | $0.11 | 6.6 |
| 82 | Phi-3 Mini Instruct 3.8B Microsoft | 4.6 | 4.6 | — | 6.6 |
| 83 | Jamba Reasoning 3B AI21 Labs | 4.1 | 4.1 | — | 6.6 |
| 84 | Phi-4 Microsoft | 4.9 | 4.9 | $0.22 | 6.5 |
| 85 | Phi-4 Mini Instruct Microsoft | 3.0 | 3.0 | $0.00 | 6.5 |
| 86 | Granite 4.1 3B IBM | 3.2 | 3.2 | — | 6.5 |
| 87 | Magistral Medium 1.2 Mistral | 20.1 | 20.1 | $2.75 | 6.4 |
| 88 | Exaone 4.0 1.2B LG AI Research | 2.9 | 2.9 | — | 6.4 |
| 89 | Granite 4.0 H 1B IBM | 2.7 | 2.7 | — | 6.4 |
| 90 | Jamba 1.5 Mini AI21 Labs | 2.7 | 2.7 | $0.25 | 6.3 |
| 91 | Tiny Aya Global Cohere | 1.0 | 1.0 | $0.00 | 6.3 |
| 92 | Nova Pro Amazon | 7.7 | 7.7 | $1.40 | 6.1 |
| 93 | Nova 2.0 Pro Preview Amazon | 20.1 | 21.8 | $3.44 | 6.0 |
| 94 | Claude Opus 4.8 Anthropic | 55.9 | 55.7 | $10.00 | 5.5 |
| 95 | Claude Opus 4.7 Anthropic | 53.7 | 53.5 | $10.00 | 5.3 |
| 96 | Llama 3.1 Instruct 405B Meta | 8.5 | 8.5 | $3.69 | 4.8 |
| 97 | Jamba 1.7 Large AI21 Labs | 5.3 | 5.3 | $3.50 | 4.6 |
| 98 | GPT-5.5 OpenAI | 54.8 | 54.8 | $11.25 | 4.6 |
| 99 | Jamba 1.5 Large AI21 Labs | 5.1 | 5.1 | $3.50 | 4.6 |
| 100 | Jamba 1.6 Large AI21 Labs | 5.0 | 5.0 | $3.50 | 4.6 |
| 101 | Nova Premier Amazon | 12.7 | 12.7 | $5.00 | 4.4 |
| 102 | Claude Opus 4.6 Anthropic | 43.7 | 43.7 | $10.00 | 4.4 |
| 103 | Command-R+ Cohere | 3.0 | 3.0 | $6.00 | 3.0 |
| 104 | Grok 4 xAI | 33.3 | 33.3 | $11.00 | 2.8 |
| 105 | Claude Fable 5 Anthropic | 60.1 | 59.9 | $20.00 | 0.0 |
Get the monthly PM AI rankings
New frontier models ship constantly. Get the updated PM Index rankings — and which model to use for which PM task — once a month. No spam, unsubscribe anytime.
Best AI model by PM use case
The right model depends on the job. These rankings reweight the benchmarks for specific product-management tasks.
How the PM Index works
Generic AI leaderboards optimize for coding or competitive math. The PM Index reweights Artificial Analysis benchmarks for the work product managers actually do — emphasizing general reasoning and multi-step competence, and de-emphasizing pure coding. It sits on the same 0–100 scale as the underlying indices so you can compare directly. Price shown is the AA-convention blended rate (3 parts input to 1 part output) per 1M tokens.