Best AI models for coding

The top AI models for real-world coding, ranked by the Artificial Analysis Coding Index — a composite score across coding benchmarks. Every one is available on anyAInow from a single balance; pay only for what you use, no subscription.

#ModelAA Coding Index
1Claude Opus 5
AnthropicFlagship·from $40 / 1M
78Try →
2GPT-5.6 Sol
OpenAIFlagship·from $50 / 1M
77.4Try →
3GPT-5.6 Terra
OpenAIBalanced·from $30 / 1M
76.7Try →
4Claude Fable 5
AnthropicFlagship·from $60 / 1M
76.5Try →
5Kimi K3
Moonshot (Kimi)Flagship·from $30 / 1M
76.2Try →
6GPT-5.5
OpenAIFlagship·from $50 / 1M
74.9Try →
7Claude Opus 4.8
AnthropicFlagship·from $30 / 1M
74.3Try →
8Grok 4.5
xAIFlagship·from $20 / 1M
72.4Try →
9Claude Sonnet 5
AnthropicBalanced·from $20 / 1M
71.5Try →
10GPT-5.6 Luna
OpenAIFast·from $10 / 1M
71.4Try →
11GPT-5.4
OpenAIBalanced·from $30 / 1M
71.1Try →
12Gemini 3.5 Flash
GoogleBalanced·from $20 / 1M
70.1Try →
13Gemini 3.1 Pro
GoogleFlagship·from $20 / 1M
68.8Try →
14GLM 5.2
Z.ai (GLM)Open·from $10 / 1M
68.8Try →
15Qwen 3.7 Max
QwenFlagship·from $10 / 1M
66Try →
16Claude Sonnet 4.6
AnthropicBalanced·from $30 / 1M
63Try →
17Kimi K2
Moonshot (Kimi)Balanced·from $10 / 1M
61.8Try →
18MiMo V2.5
Xiaomi (MiMo)Open·from $5 / 1M
60.2Try →
19DeepSeek V4 Pro
DeepSeekOpen·from $10 / 1M
59.4Try →
20MiniMax M3
MiniMaxOpen·from $10 / 1M
58.6Try →
21DeepSeek V4 Flash
DeepSeekFast·from $5 / 1M
56.2Try →
22GPT-5.4 mini
OpenAIFast·from $10 / 1M
56.1Try →
23Qwen 3.7 Plus
QwenOpen·from $10 / 1M
55.9Try →
24Qwen 3.6 Plus
QwenOpen·from $10 / 1M
54.5Try →
25MiniMax M2.7
MiniMaxOpen·from $10 / 1M
52.6Try →
26Grok Build
xAIFast·from $10 / 1M
51.5Try →
27Mistral Medium 3.5
MistralBalanced·from $20 / 1M
46.9Try →
28Claude Haiku 4.5
AnthropicFast·from $10 / 1M
43.9Try →
29Grok 4.3
xAIFlagship·from $10 / 1M
42.2Try →
30Gemini 3.1 Flash Lite
GoogleFast·from $10 / 1M
34.7Try →
31Gemini 2.5 Pro
GoogleFlagship·from $20 / 1M
33.3Try →
32Mistral Small 4
MistralFast·from $5 / 1M
26.6Try →
33Mistral Large 3
MistralFlagship·from $10 / 1M
20.1Try →
34Llama 4 Maverick
Meta LlamaBalanced·from $10 / 1M
16.3Try →
35Llama 3.3 70B
Meta LlamaFree·Free
11.9Try →
36Gemma 3 27B
GoogleOpen·from $10 / 1M
10.1Try →
37Llama 4 Scout
Meta LlamaOpen·from $5 / 1M
8.2Try →

Benchmarks via artificialanalysis.ai, as of 13 Aug 2026. Figures are publisher-reported where not independently verified. Prices shown are the anyAInow pay-as-you-go floor (per 1M tokens); free-tier models deduct nothing.

More rankings