Updated 4 November 2026 · Artificial Analysis

The best value AI model

$0.01$0.10$1.00$10$1000102030405060Nemotron 3.5 Lightning · 6 · $0.10Granite 4.2 8B · 5 · $0.11MiMo-V2.6-Flash · 27 · $0.17Claude Haiku 5.5 (Max) · 37 · $0.20GPT-6 Luna (Max) · 28 · $0.20GLM-5.3-Flash · 28 · $0.24MiniMax-M3.1-Flash · 29 · $0.26Mistral Small 4 (Reasoning) · 4 · $0.26Nemotron 3.5 Super VL · 14 · $0.28Ling 3.1 Flash 1019 · 30 · $0.45GPT-5.6 Luna (Max) · 24 · $0.45DeepSeek V4.1 Flash 1017 (Max) · 33 · $0.53MiniMax-M3 · 19 · $0.53Inkling Small · 13 · $0.53Muse Glimmer (High) · 6 · $0.53MiMo-V2.6-Pro · 30 · $0.54Gemini 3.1 Flash-Lite · 6 · $0.56Beam 1.0 · 18 · $0.60Gemini 3.5 Flash-Lite · 12 · $0.85Mistral Large 4 (High) · 26 · $1.05Mistral Large 4 Preview · 25 · $1.05Qwen3.8 27B (Xhigh) · 20 · $1.13Nemotron 3 Ultra 550B A55B (Reasoning) · 12 · $1.13Step 5 Preview · 29 · $1.43Gemini 3.8 Flash (High) · 31 · $1.50Inkling (Xhigh) · 13 · $1.76DeepSeek V4 Pro 0813 (Max) · 23 · $1.98GPT-6.1 Terra (Max) · 51 · $2.00Muse Spark 1.3 (Max) · 37 · $2.00Muse Spark Open (Xhigh) · 25 · $2.00Claude 4.5 Haiku (Reasoning) · 7 · $2.00GLM-5.3 (Max) · 32 · $2.15Grok 4.7 (Xhigh) · 36 · $3.00Qwen3.8 Max (0902) · 34 · $3.00Grok 4.6 (Xhigh) · 31 · $3.00Mistral Medium 3.5 · 5 · $3.00GPT-6.1 Sol (Max) · 55 · $4.00Claude Sonnet 5.5 (Max, Default Fallback) · 55 · $4.00Kimi K3.1 Pro · 47 · $4.00GPT-6 Sol (Max) · 44 · $4.00GPT-5.6 Terra (Max) · 32 · $4.50Kimi K3 (Low) · 17 · $6.00Gemini 4 Carbon (High) · 60 · $8.00Claude Opus 5.5 (Max, Default Fallback) · 59 · $8.00Gemini 4 Argon (High) · 50 · $8.00Muse Wonder (Max) · 44 · $8.00GPT-6 Astra (Max) · 57 · $20Claude Fable 5.1 (Xhigh, Default Fallback) · 50 · $20Nemotron 3.5 LightningMiMo-V2.6-FlashClaude Haiku 5.5GPT-6.1 TerraGPT-6.1 Solprice per 1M tokens (log scale) →↑ intelligence
Orange: the value frontier, models no cheaper model beats. Grey: everything else.

Best model by budget

Budget per 1M tokensSmartest modelIntelligencePrice
Under $1Claude Haiku 5.5 Anthropic37$0.20
Under $5GPT-6.1 Sol OpenAI55$4.00
Under $15–50Gemini 4 Carbon Google60$8.00

The value frontier

#ModelIntelligencePrice / 1M tokensInput / 1MOutput / 1M
1Gemini 4 Carbon Google60$8.00$4.00$20
2GPT-6.1 Sol OpenAI55$4.00$2.00$10
3GPT-6.1 Terra OpenAI51$2.00$1.00$5.00
4Claude Haiku 5.5 Anthropic37$0.20$0.10$0.50
5MiMo-V2.6-Flash Xiaomi · open weights27$0.17$0.14$0.28
6Nemotron 3.5 Lightning NVIDIA · open weights6$0.10$0.06$0.20

A model is on the value frontier when no cheaper model scores higher on the Artificial Analysis Intelligence Index. Price is the blended price per million tokens (3 parts input to 1 part output), the cheapest hosted rate tracked by Artificial Analysis.

Data: Artificial Analysis (intelligence and coding indexes, prices, speed) and OpenRouter (open weights, context windows). Updated 4 November 2026; refreshed every 6 hours.

More rankings

Watch the live race